The Frontiers of Society,

Science and Technology

Issue 04

2018

Call for Papers

The Frontiers of society, Science and Technology (TFSST, ISSN 2616-7433) is a scholarly peer-reviewed

international scientific journal published Bimonthly for researchers, developers, technical managers, and

educators in the engineering and society field. It provides a high profile, leading edge forum for academic

researchers, industrial professionals, engineers, consultants, managers, educators and policy makers

working in the field to contribute and disseminate innovative new work on all the areas of society, science

and technology.

Submission Guidelines

Papers are accepted either through the online submission tool or email at tfsstCFP@hotmail.com. For those

unable to use the Online System and are emailing, please include a cover letter with author details,

university or country information, and any other author related data.

⚫ .docx/.pdf/Latex PDF file formats are accepted.

⚫ The covering email should clearly state the name, address and affiliation of the corresponding author.

Each paper will be provided with a paper ID for further reference.

⚫ Please ensure all articles are proofread and checked for accuracy prior to submission. Changes cannot

be made during the review process. If accepted, minor revisions may be made prior to final

submission.

Formatting Guidelines

⚫ Papers must be in English of up to 4-5 pages.

⚫ Paper content must be original and topics of interests are not limited.

⚫ Authors are required to ensure accuracy of quotations, citations, diagrams, maps, and tables.

⚫ Figures and tables need to be placed where they are to appear in the text and must be clear and easy to

view.

⚫ Papers must follow format according to the downloadable Word or Latex template.

Review Process

The acceptance decisions will take into account paper novelty, technical depth, elegance, practical or

theoretic impact, and presentation.

⚫ Original: the paper explores a new idea, project or issue; discusses existing research with promise of

new insight, discusses new research; or presents new ways of considering existing information.

⚫ Significant: the paper raises and discusses issues important to improving the effectiveness and/or

sustainability of open education efforts, and its contents can be broadly disseminated and understood

⚫ Quality: claims are supported by sufficient data; claims draw upon relevant literature; and limitations

are described honestly.

The review process may take approximately three months to be completed. Should authors be requested by

the editor to revise the text, the revised version should be submitted within 1 month. Authors who need

more time are kindly requested to contact the Editor. The Editor reserves the right to reject a paper if it does

not meet the aims and scope of the journal, it is not technically sound, it is not revised satisfactorily, or if it

is inadequate in presentation.

Content

Research on Improved Genetic Algorithm for Virus Intrusion Detection Model

R. Vijayanand, D. Devaraj, B. Kannapiran P1

Differential Drive and Steering Control Technology of Automatic Navigation

Handling Tool Based on PLC

Takao Kobayashi, Etsuo Katsuyama, Hideki Sugiura P6

Application of Ozone MBBR Process in Refinery Wastewater Treatment

Henrik Rasmus Andersen, Kamilla M. S. Kaarshol, Ravi Kumar Chhetri P12

Association Rules Redundancy Processing Algorithm Based on Hypergraph in

Data Mining

Ankit Agrawal, Alok Choudhary P18

Research and review on the knot of Innovation and Entrepreneurship Training

Program project in Leshan Normal University in November 2018

Zhao Li xin P27

The Frontiers of Society, Science and Technology

1

Research on Improved Genetic Algorithm for

Virus Intrusion Detection Model

R. Vijayanand1,*, D. Devaraj2, B. Kannapiran3

1 Shanghai Advanced Research Institute, Chinese Academy of Sciences (CAS), Shanghai, 201210, China

2 Department of Electrical and Electronics Engineering, Kalasalingam University, Srivilliputur, Tamilnadu, India

University of the CAS, Beijing 100039, China

3 Department of Instrumentation and Control Engineering, Kalasalingam University, Srivilliputur, Tamilnadu,

India

*E-mail: rkvijayanand@gmail.com

Abstract: The purpose of this study is to solve the

computer security problems. Based on the traditional

security model, the hidden dangers of computer

security and the advantages and disadvantages of

various intrusion detection techniques are analyzed. A

computer intrusion detection system based on an

improved genetic algorithm is designed. The results

show that the optimized genetic algorithm can

improve the efficiency of intrusion detection and

reduce the false alarm rate. Therefore, we conclude

that the application of genetic algorithm in intrusion

detection has important theoretical and practical

significance.

Keywords: improved genetic algorithm; intrusion

detection model; traditional security model; computer virus

1. INTRODUCTION

With the rapid development of computer technology

and communication technology, the application of

computer is becoming more and more popular [3]. At

the same time, the computer security problem

becomes more and more complex. In the late 1960s,

people realized the vulnerability of computer system

and began to study its security [2]. In the1980s, the

performance of computers has been greatly improved.

All the isolated computer systems are connected to

form a network to realize mutual communication and

resource sharing [5]. Especially since the 1990s, the

rapid development of the Internet has greatly

promoted the social informatization [1]. Based on the

computer network environment, it has realized a

variety of applications such as electronic banking,

e-commerce, e-government, e-household, financial

network, manufacturing resource management and

network virtual community [4]. However, the

network also provides an opportunity for illegal acts

of information. The storage, processing and

transmission of information is attacked by the

intruder or the virus, which results in the paralysis of

the system or the loss of important data. For example,

hackers break new industrial security mechanisms,

and millions of computers are paralyzed for viruses in

hours [6]. In a word, computer security becomes

more and more important. It is related to business

interests, personal privacy and national secrets.

Therefore, the computer security issues need to be

resolved urgently. It is of great significance for

China's aerospace manufacturing industry [8].

From the initial experimental research to today's

commercial products, intrusion detection system (IDS)

has been developed for more than 20 years. In 1980,

Anderson first proposed audit tracking, which could

be used to monitor intrusion threats [9]. Although the

importance of the idea was not understood at the time,

his report was considered groundbreaking work for

intrusion detection. From 1984 to 1986, Dorothy

Denning and Peter Neumann jointly proposed a

real-time intrusion detection system model, which

was named IDES (Intrusion Detection Expert

System). It is a host based intrusion detection system,

which uses statistical methods to detect user

abnormal behavior. Their hypothesis is the basis of

intrusion detection research. In 1987, Dorothy

Denning introduced a generic intrusion detection

model [10]. The model consists of three components:

event generator, activity recorder and rule set. The

event generator is part of the model that provides

system activity information. The activity recorder

holds the status of the system or network in the

monitoring. The rule set is an ordinary check event

for checking events and status data. In 1980s,

inspired by Anderson and IDES, a large number of

IDS systems emerged [11]. However, until the 1990s,

the commercial IDS appeared. In foreign countries,

commercial intrusion detection systems include ISS's

Secure Secure, Nai Cybercop, SVC Net Prowler,

CA's Session Wall-3, Cisco's Secure IDS and IBM's

IERS (Internet Emergency Response Service). In

China, there are few network intrusion software [13].

It mainly includes intrusion detection systems of

Netpower, hacker intrusion detection and early

warning system. There are still a big gap with foreign.

In 1994, Biswanath Mukherjee made a complete

review and analysis of previous IDS studies. He

The Frontiers of Society, Science and Technology

2

analyzed and commented on various IDS prototype

systems. After 1995, there were many different new

IDS research methods, especially intelligent IDS,

which included neural networks, fuzzy recognition,

genetic algorithms, immune systems, and data mining

[12]. These methods are currently in the theoretical

research stage. At present, most commercial intrusion

detection systems are similar in principle to virus

detection. They have a certain size and number of

intrusion feature pattern libraries, which can be

updated regularly [7].

2. MATERIALS AND METHODS

(1) The concept of genetic algorithm

Genetic algorithm (GA) is a global search algorithm,

which randomly generates a population. Then. it

simulates the process of natural evolution in a better

direction, and evaluates the merits of individuals

through fitness. In each generation, the superior

individual produces offspring by genetic

manipulation, and the inferior individual is

eliminated. Individuals in a population are usually

composed of a string of binary or real numbers of

length. An individual is a solution to the problem, and

genetic algorithms are widely used to solve

optimization problems.

The genetic algorithm was first proposed by

American professor John. Holland of the University

of michigan. It originated from the study of adaptive

behavior of natural and artificial systems in the 60s. It

is a global optimization random search algorithm,

which imitates the principles of natural biology,

heredity and evolution. It uses bit string coding

technique to generate initial population for the

problem. Then, the fitness of the population is

evaluated, followed by selection, crossover, and

mutation operations. A selection strategy based on

fitness ratio is used to select individuals in the current

population, and crossover and mutation are used to

produce the next generation. This generation goes on

until it meets the expected conditions. The genetic

algorithm can simultaneously search multiple regions

in the solution space by using the method of

population organization, and it is especially suitable

for large-scale parallel processing. Genetic algorithm

has the characteristics of self-organization,

self-adaptation and self-learning. The natural

selection of the fittest and the simple genetic

operations make the computation free from the

constraints of the search space (e.g., differentiable,

continuous, unimodal) and the absence of additional

supporting information (such as guidance). Therefore,

the genetic algorithm not only can achieve high

efficiency, but also has a simple, easy to operate and

common features. At present, with the development

of computer technology, genetic algorithm has been

paid more and more attention, and has been

successfully applied in machine learning, pattern

recognition, image processing, combinatorial

optimization, VLSI design and optimization control.

Genetic algorithm is not a simple random comparison

search algorithm. By assessing the fitness of

chromosomes and the role of genes in chromosomes,

it effectively uses existing information to guide the

search for the most promising states to improve the

quality of the optimization. The flow chart of simple

genetic algorithm is shown in Figure 1.

Figure 1. The flow chart of simple genetic

algorithm

(2) Improvement of genetic algorithm

In the genetic algorithm, the generation of new

generation population mainly depends on the

stochastic crossover, recombination and mutation of

the previous generation. Therefore, it takes a great

deal of cost to obtain the optimal solution. Therefore,

genetic algorithm has the disadvantages of premature

maturity and weak local search ability, which

seriously affect the efficiency of genetic algorithm.

Nowadays, the commonly used methods of

improvement can be broadly divided into the

following categories. First, genetic operators are

improved appropriately, including selection,

crossover and mutation, etc. in addition, in order to

improve the performance of the algorithm, a number

of high-level genetic operations have been developed

and applied. Second, combining genetic algorithm

with traditional heuristic search techniques based on

problem knowledge (such as hill climbing, simulated

annealing, tabu search, etc.), a framework of hybrid

search algorithm based on genetic algorithm is

constructed. Third, the genetic algorithm is

parallelized. At present, the most basic parallel

schemes are synchronous master-slave, asynchronous,

synchronous and network. Fourth, the dynamic

adaptive technology is used to adjust the algorithm

control parameters and coding granularity in the

evolution process. The last one is an organic

combination of several previous approaches.

This paper mainly studies the use of genetic

algorithm to train the neural network, optimize the

network connection right, and use it for intrusion

detection. Due to the shortcomings of the local search

The Frontiers of Society, Science and Technology

3

ability of the basic genetic algorithm, it is necessary

to improve the genetic algorithm.

3. RESULTS

(1) Implementation of improved genetic algorithm

The main idea of genetic algorithm to optimize the

weights of BP neural network is as follows. First of

all, the genetic algorithm is used to optimize the

initial weight (threshold) distribution, and a better

search space is found in the solution space. Then, BP

algorithm is used to search the optimal solution in

this smaller solution space. The improved genetic

algorithm can optimize the weights of neural

networks and prevent the search from being trapped

into local minima.

The global characteristics of genetic algorithm and

the fast and efficient characteristics of neural

networks in local optimization are comprehensively

utilized. A local improved GABP algorithm is used to

train the neural network. A locally improved genetic

algorithm is used to optimize the weights of neural

networks. The adaptive learning rate momentum

gradient descent algorithm is used to train the neural

network, and the fitness function is calculated.

Finally, the output of neural network is computed by

the weight corresponding to the maximum fitness

function. The algorithm flow is shown in Figure 2.

Figure 2. The algorithm flowchart

(2) Detector sets

The generation of the detector is calculated and

analyzed in a probabilistic manner. If Nx is used to

represent the size of the set X, then: NR0: the number

of candidate detectors, that is, the size of the detector

set prior to review; NR: the number of valid detectors,

that is, the size of the set of detectors after the review;

NS: the number of strings in the "own" collection; PM:

the probability of matching between any two random

strings; Pf: false dismissal probability,that is,

Pf=(1-PM)NR; f: the probability that any random string

does not match the NS string in the "own" set, that is,

f=(1-PM)NS.

Because f=(1-PM) NS, we can get

)MP1(In•SN=Inf － . When the PM is small enough,

according to the Taylor expansion:

SNMPe•

MPfPnI

=0RN﹣

(1)

This formula can estimate the number of candidate

detectors NR0. NR0 is the function of Pf, NS and PM.

The minimum set of detectors is obtained by the

formula (1). Then, by calculating the minimum, we

can get:

SN1

=MP (2)

Therefore, when matching rules SN

1=MP , there is

the minimum value of NR0. At this point, the required

set of candidate detectors is minimal, that is, the

number of randomly generated strings is minimized.

Formula (1) gives an estimate of NR0, which is related

to the size of the protected content NS, the required

detection reliability Pf, and the matching rule PM.

Through further analysis, we can see that the

algorithm has the following characteristics. The

formula (2) is introduced into formula (1), and the

function relation between NR0 and NS is obtained:

SN•)fInP•e-(=0RN . When choosing a suitable

detection probability Pf, )fInP•e-( is a constant.

Therefore, NR0 is proportional to NS. Figure 3 shows

the minimum NR0 dynamic results for each different

NS value. It can also be seen from equation (1) that

the size NR0 of the detector set is related to the

detection capability Pf of the detection system.

The elements of the detector set should have better

distribution characteristics, so that neither of the two

detectors will match the same "non-self" string. The

string space that the detector set R is covered in the

collection U is: UN×RN×MP . For a given failure

probability Pf, it requires at least

)SN-UN(×)fP-1( detector in order to cover the

"non-self" string space )SN-UN(×)fP-1( that can

be detected. Therefore, we can get: )SN-UN(×)fP-1(≥UN×RN×MP

That is,

MP

)fP-1(≥RN )SNUN( ＞ (3)

If Pf and PM are fixed, then NR0 grows exponentially

with NS. Therefore, NR0 is very sensitive to NS, that is,

NS has changed a little, it will lead to a large set of

candidate detector changes. For this index factor, we

can consider from both positive and negative aspects:

the negative side is that it will increase the

computational cost of the algorithm; its advantage is

that it makes it harder for the intruders to cover up

their malicious behavior. For example, if a collection

of detectors is generated (possibly using a

supercomputer), an attempt to mask its malicious

behavior is difficult to succeed if the malicious agent

The Frontiers of Society, Science and Technology

4

wants to modify the "own" collection, and then

modify the set of detectors accordingly.

(3) Algorithm test

The system runs on Red Hat Linux 9.0 and

windows2000 server platform. The simulation

experiments use MATLAB 7.4.0. The trial CPU is

AMD Athlon XP 1700+, memory 1G and 120G hard

drive. The database uses SQL 2000 Server. The

actual network data packet is used as the data source,

and the experimental results of GABPIDS in the

simulation are obtained, and the detection rate, false

alarm rate and false negative rate are calculated. This

experiment uses the more authoritative KDD CUP 99

data set for intrusion detection.

The data set contains normal data and a variety of

anomaly data, involving a total of four major

categories of 22 categories of intrusion attacks: DOS

(denial of service attacks); R2L (remote

non-authorized access attacks); U2R (unauthorized

access to super user rights attacks); PROBE

(vulnerability detection and scanning attacks). The

kddcup.data.corrected file provided by the dataset

contains 4,898,431 raw data, SQL2000 delete

duplicate records, get 1,074,992 valid data, the

distribution is: Normal (812,814), DOS (247,267),

R2L (999), U2R (52), PROBE (13,860).

The data used in the experiment are pretreated by

PCA technology. After the 41-dimensional data is

processed, the data dimension is reduced to 7, which

greatly reduces the dimension of the input data of the

neural network and the training and testing time of

the neural network, and improves the system

effectiveness.

In order to explain the performance of genetic

algorithm to optimize the design of neural network

more clearly, this paper compares BP algorithm with

GABP algorithm. Each algorithm was trained 100

times and compared with the same test data. The

convergence times, convergence speed and average

detection results of each algorithm are shown in

Table 1 and Table 2.

Table 1. The convergence effect statistical

The number of convergence in 100 training sessions The convergence speed

Neural network algorithm 66 Slower

Improved genetic algorithm 100 From fast to slow

Table 2. Detection effect statistical

Attack type Training /

test

samples

Neural network algorithm Improved genetic algorithm

Average false

alarm rate

Average

detection rate

Average

false alarm

rate

Average

detection rate

neptune (DOS) 5000/1000 2.5% 95.6% 1.6% 99.2%

smurf (DOS) 5000/1000 4.8 % 92.1% 3.6% 96.5%

ipsweep (PROBE) 5000/1000 5.1% 89.1% 4% 95%

portsweep (PROBE) 5000/1000 7.3% 87.5% 6.8% 94.8%

satan (PROBE) 5000/1000 8.1% 89.3% 6.5% 95.3%

Mixed attack 10000/2000 12.3% 73.5% 10% 85.3%

Mixed attack, test data for

new attacks

10000/2000 12.8% 60.5% 10.9% 74.3%

As can be seen from Table 1 and Table 2, compared

with the neural network algorithm, the optimized

genetic algorithm not only converges fast and

converges easily, but also has higher detection rate

for various attacks and lower false positive rate. At

the same time, compared with other current methods,

the detection rate has also improved significantly. It

also proves that the performance of the optimized

network is improved obviously after the improved

genetic algorithm.

The improved genetic algorithm is used to optimize

the neural network algorithm, and the simulation test

is carried out on the MATLAB platform. Through the

preprocessing of real network packets, the input data

of the improved genetic algorithm is obtained. The

reduction of dimension is reduced by introducing

PCA technology, and the structure of neural network

is simplified. Neural networks and improved genetic

algorithms are used to test the same data. It is found

that the improved genetic algorithm is easier to

converge, and the convergence speed is faster. It has

high classification and recognition rate for normal

data. For the data of mixed attack and new attack,

although the detection rate of the improved genetic

algorithm has been improved significantly, it is still

around 75%. For single attack data recognition,

especially for Neptune (DOS) attack data, the

detection rate is between 98%~100%, and the

recognition effect is very good.

4. CONCLUSION

Combined with the actual data of computer network,

an intrusion detection model based on genetic neural

network is proposed. Through the data acquisition,

preprocessing and feature extraction, the model

transforms the computer network data into neural

network data. In order to improve the deficiency of

the basic genetic algorithm, the genetic algorithm is

optimized by the fitness value calibration and the

expected value selection method. In this way, the

improved genetic algorithm is more suitable for

intrusion detection system. Through analysis and

research, the following conclusions are obtained.

The Frontiers of Society, Science and Technology

5

Compared with the traditional genetic algorithm, the

improved genetic algorithm has been improved in

terms of system efficiency, detection rate and

omission rate. At the same time, it improves the

ability to detect unknown attacks to a certain extent.

REFERENCE

[1] Ahmad, I., Hussain, M., Alghamdi, A., & Alelaiwi,

A. (2014). Enhancing SVM performance in intrusion

detection using optimal feature subset selection based

on genetic principal components. Neural Computing

and Applications, 24(7-8), 1671-1682.

[2] Bhattacharjee, P. S., Fujail, A. K. M., & Begum, S.

A. (2017). Intrusion Detection System for NSL-KDD

Data Set using Vectorised Fitness Function in Genetic

Algorithm. Advances in Computational Sciences and

Technology, 10(2), 235-246.

[3] Chen, M. H., Chang, P. C., & Wu, J. L. (2016). A

population-based incremental learning approach with

artificial immune system for network intrusion

detection. Engineering Applications of Artificial

Intelligence, 51, 171-181.

[4] Dastanpour, A., Ibrahim, S., & Mashinchi, R.

(2014). Using genetic algorithm to supporting

artificial neural network for intrusion detection

system. In The international conference on computer

security and digital investigation (ComSec2014) (pp.

1-13). The Society of Digital Information and

Wireless Communication.

[5] Elhag, S., Fernández, A., Bawakid, A.,

Alshomrani, S., & Herrera, F. (2015). On the

combination of genetic fuzzy systems and pairwise

learning for improving detection rates on Intrusion

Detection Systems. Expert Systems with Applications,

42(1), 193-202.

[6] Ganapathy, S., Kulothungan, K., Muthurajkumar,

S., Vijayalakshmi, M., Yogesh, P., & Kannan, A.

(2013). Intelligent feature selection and classification

techniques for intrusion detection in networks: a

survey. EURASIP Journal on Wireless

Communications and Networking, 2013(1), 271.

[7] Hosseinpour, F., Amoli, P. V., Farahnakian, F.,

Plosila, J., & Hämäläinen, T. (2014). Artificial

immune system based intrusion detection: Innate

immunity using an unsupervised learning approach.

International Journal of Digital Content Technology

and its Applications, 8(5), 1.

[8] Jahromy, B. M., Honarvar, A. R., Saif, M., &

Jahromy, M. A. M. (2016). A New Method for

Detecting Network Intrusion by Using a Combination

of Genetic Algorithm and Support Vector Machine

Classifier. Journal of Engineering and Applied

Sciences, 100(4), 810-815.

[9] Mukesh, K. G., Khanna, H. P., & Velvizhi, R. V.

(2015). An anamoly based Intrusion Detection

System for mobile ad-hoc networks using genetic

algorithm based support vector machine. Advances in

Natural and Applied Sciences, 9(12), 40-45.

[10] Moorthy, V. P., & Kumar, S. A. (2014).

MCKELM-IDS: efficient feature transformation &

optimal feature subset selection based intrusion

detection approach using MCKELM. Advances in

Natural and Applied Sciences, 8(19), 88-100.

[11] Shamshirband, S., Amini, A., Anuar, N. B., Kiah,

M. L. M., Teh, Y. W., & Furnell, S. (2014). D-FICCA:

A density-based fuzzy imperialist competitive

clustering algorithm for intrusion detection in

wireless sensor networks. Measurement, 55, 212-226.

[12] Singh, T., Verma, S., Kulshrestha, V., & Katiyar,

S. (2016, March). Intrusion Detection System Using

Genetic Algorithm for Cloud. In Proceedings of the

Second International Conference on Information and

Communication Technology for Competitive

Strategies (p. 115). ACM.

[13] Swaminathan, A., Vivekanandan, P., & Sivajothi,

E. (2016). Anomaly Detection Model Based on

Multivariate Correlation Analysis Technique to

Detect Covert Communication in Wireless Sensor

Network. Journal of Computational and Theoretical

Nanoscience, 13(8), 5281-5287.

The Frontiers of Society, Science and Technology

6

Differential Drive and Steering Control

Technology of Automatic Navigation

Handling Tool Based on PLC

Takao Kobayashi1,*, Etsuo Katsuyama2, Hideki Sugiura1

1 Mechanical Engineering Dept. 2, Toyota Central R&D Laps., Inc., Aichi, Japan

2 Classis Control Development Div., Toyota Motor Corporation, Shizuoka, Japan

*E-mail: e1545@mosk.tytlabs.co.jp

Abstract: In order to improve the differential drive

and steering control technology, the differential drive

and steering control technique for automatic

navigation based on power line communication (PLC)

is designed. The steering of the automatic guided

vehicle (AGV) is divided into steering and

differential steering. Because of the different steering

modes, the AGV will cause different errors when

moving along the desired trajectory in the path

trajectory transformation. The AGV locus consists of

straight and arc line. Among them, the steering

trajectory of AGV is not complex because of the lack

of spin adaptation, so its trajectory transition is

complicated. Taking the self-developed single wheel

drive and steering AGV as the object of study, the

AGV trajectory planning program and motion

program are developed by using Matlab and VC++.

The error of AGV in trajectory transformation is

analyzed theoretically, and it is proved that the error

caused by acceleration of motor can be compensated.

At the same time, combined with the development of

AGV control card PMAC, the program of motion is

discussed. In addition, visual navigation has become

the main direction of navigation technology because

of its large amount of information and flexibility. In

order to improve the robustness and accuracy of

landmark detection, the effects of image denoising,

edge detection and other techniques in navigation are

discussed in turn. For the identification of path

trajectory, genetic algorithm is introduced to improve

the accuracy and time truth of navigation. Finally, the

experiment proves the feasibility of the theory, and

also gives the method to reduce the error of vision

navigation. It provides a good foundation for

improving the accuracy of visual navigation.

Keywords: single wheel drive / steering AGV, turn

error, visual navigation, genetic algorithm, PLC

1. INTRODUCTION

PLC (Power Carrier Communication) is a

communication technology that couples data signals

to power lines and uses power lines as propagation

media. With the development of DSP and digital

coding technology, PLC technology will be used

more widely. As an important means of

communication, PLC technology can replace the

special communication network through the existing

power line network. It is easy to implement the

lighting control automation, security monitoring,

remote meter reading and so on. According to the

bandwidth size, PLC technology can be divided into

narrowband PLC technology (9-500kHz) and

broadband PLC technology (2-34, MHz).

Narrowband PLC technology is often used in

situations where data transmission rate is not high,

such as power grid load control, automatic meter

reading system and street light control. Broadband

PLC is commonly used in high-speed data

transmission occasions, such as large file transceiver,

voice calls, video calls and other high-speed data

transmission services. As one of the greatest

inventions of mankind in twentieth century, robot has

made great progress after more than 50 years'

development, especially mobile robot. Because of its

unique advantages, it has been widely used in all

walks of life. As a member of mobile robot, it has the

characteristics of good flexibility, automation and

high intelligence [1]. The path can be changed

according to the storage and production process

changes. It has realized the automation of loading and

unloading of goods and materials, and has become

the main logistics tool in FMS.

According to the definition of American Logistics

Association, AGV refers to a vehicle equipped with

an electromagnetic or optical automatic guidance

device. It can travel along a prescribed path. At the

same time, it also has car programming and parking

selection device, security protection and various load

removal functions. The AGV is an unmanned

automatic transport vehicle with a contactless

guidance device and an independent addressing

system. It is powered by batteries, and it is an

automatic material handling equipment developed in

50 and 60s. The utility model can be used for

transportation of goods in docks, airports, hospitals,

shops, etc., and can also be used as a movable

assembly platform for assembly lines. In particular,

there is an urgent need for dangerous workplaces

The Frontiers of Society, Science and Technology

7

where staff cannot operate [2].

Through theoretical analysis, the method of solving

this problem for single wheel AGV is discussed. The

attitude of black guiding belt is recognized by CCD

camera, and the AGV motion is guided by the

acquired data. Since the curvature of the black band

is guided to guide the AGV, the single wheel AGV

can be satisfied as long as the rudder angle is

controlled. But for two-wheel differential drive AGV,

the control is more complicated than single wheel

drive. Therefore, the visual navigation of two

wheeled AGV is mainly discussed.

2. LITERATURE REVIEW

The first AGV in the world was developed by Barrett

electronics company in America in 1950s. Prior to the

middle of the 80s, most of the AGV was fixed wired

guidance, limiting the restructuring and layout of its

production. With the rapid development of modern

industry, the level of AGV technology is also

increasing. AGV is equipped with on-board

computers, communication devices and materials

handling devices, and the overall complexity and

automation are greatly improved [3]. The United

States government set up the National Automatic

Highway Alliance (NAHSC) in 1995 to study the

feasibility of developing an automatic guided vehicle

and to promote the practicability of AGV. Generally

speaking, the development of mobile robots in

foreign countries basically follows two lines. Some

universities and research institutions in Europe follow

the academic line that robots are used as a vehicle for

research into artificial intelligence. Since artificial

intelligence and other intelligent technologies are far

from expected, this kind of research is mostly in the

experimental stage.

In recent years, the research and development of

automatic guided vehicle (AGV) in our country

shows a thriving scene, which is mainly related to the

logistics heat in recent years. Beijing Hoisting

Machinery Research Institute, China Postal Science

Planning Institute, CAS Shenyang Automation

Institute, Dalian composite machine tool research

institute and other research institutes are the earliest

units involved in the development of AGV. Some of

them have been produced in small quantities. AGV,

the company that produces the real scale, is the

Kunming marine equipment company [4]. In addition,

Shenyang SIASUN, robotic company also developed

a variety of AGV. Tsinghua University has developed

a THMR_V automatic guided vehicle, has realized

the structured Lane under the environment of

automatic tracking, quasi structured environment

road tracking under complex environment, obstacle

avoidance, telepresence driving and other functions.

National University of Defense Technology has

developed automatic guided vehicle CITAVT series.

The CITAVT_ IV vision autonomous navigation

vehicle also implements autonomous navigation on

the highway.

3. METHODS

(1) The composition and working principle of AGV

Because of its wide range of applications, the specific

purpose and working conditions of AGV vary greatly.

Therefore, it also has many kinds. But functionally, a

variety of AGV should have the following

subsystems:

Drive / boot system: The driving system and the

guidance system of AGV are inseparable, and they

are related to the chassis structure adopted. The

design of the chassis structure ensures that the AGV

can move freely and flexibly under the control of the

drive / guidance system. The commonly used is

three-wheel chassis dual rear wheel differential drive

/ steering, three-wheel chassis single-wheel drive /

steering, four-wheel dual rear wheel differential drive

/ steering, four-wheel drive dual-wheel drive /

steering, differential drive / steering, six-wheel drive

dual-wheel drive / steering. Depending on the

guidance system, the AGV can automatically run

along the ground lead wire or the ground reflector or

even any designated route [5].

Communication system: Several AGV can form an

automated guided transport system (AGVS) together.

When the automatic guided vehicle system in

multiple car works at the same time, in order to make

the car work conflict (task conflict, route conflict),

logistics automation equipment and all other car in

the system are unified by the central control computer

control [6]. The car needs to receive instructions from

the host computer for the next step, and also needs to

report its current status to the main controller, which

is done through the communication system. Fixed

line AGV can communicate by wires embedded on

the running line, while AGV can only use wireless

communication in all directions [7]. When there are

many interference sources, the communication

system must have higher reliability in order to ensure

the normal operation of the system.

(2) Main technical parameters of AGV

Rated loading capacity. It is the maximum weight

that AGV can bear.

Dead weight. It refers to the total weight added by the

AGV and the battery.

Car body size. It refers to the size of the car body, and

this size should be subject to the size of the goods

and the site.

Stopping accuracy. It refers to the number of

millimeters between the location of the AGV job and

the location of the program.

Minimum turning radius. It refers to the distance

between the instantaneous steering center distance

and the AGV longitudinal centerline when the AGV

is in the form of no-load, low speed and maximum

deflection.

Running speed. It refers to the maximum speed at

which AGV travels at rated loads.

Cell voltage. There are two specifications for battery

voltage, 24V and 48V respectively.

The Frontiers of Society, Science and Technology

8

Work cycle. It refers to the time required for AGV to

complete a cycle of work.

4. RESULTS AND DISCUSSION

AGV is a wheeled autonomous mobile robot, which

mainly completes the tasks of path planning,

positioning and obstacle avoidance. Therefore, path

planning is one of the basic steps of AGV navigation

[8]. It searches for an optimal or nearly optimal

collision free path from the start state to the target

state according to a performance index. According to

the degree of robot access to environmental

information, path planning is divided into two types.

The environment information is fully aware of the

global path planning and the environment

information is completely unknown or partially

unknown to the local path planning [9]. The global

path planning that is fully aware of the environment

information is applicable to static environments, and

the path is off-line planning. The local path planning

environment information completely or partly

unknown is detected by the sensor online on the robot

operation environment, information obstacle location,

shape and size to complete planning. The path is

planned online. This article only discusses the

planning of known paths for single round AGV.

(1) Trajectory analysis of single wheel drive / steering

AGV

According to the structure and motion analysis of

AGV, the trajectory of double wheel drive AGV can

be divided into straight line, arc and spin. When the

two wheels are the same in size and speed, the AGV

can move forward or backward in a straight line.

When the two wheels have the same speed direction,

but the size is different, the AGV can move forward

or backward in the arc motion [10]. When one wheel

does not move, the AGV will rotate at the center of a

wheel. When the two wheels move at the same speed

in the opposite direction, the AGV will spin at its

center of mass.

For a single wheel drive / steering AGV, since its

steering is controlled by the steering motor to drive

the wheel, the AGV has only two tracks of circular

and straight lines, and there is no spin. However, if

there is an angle between the front and rear two

tracks, since there is no spin, the AGV must go

through a transition arc to complete the trajectory. It

is necessary to know that AGV needs to take the

amount of arc ahead of time and the amount of

second segments that have not gone. That is, the

point between the arc and the two trajectories.

At the same time, when the AGV is taking the arc

path, the steering motor needs to control the drive to

rotate an angle. Obviously, the radius of the arc is

different. The steering motor needs to be controlled,

and the angle of rotation is different. So, it is assumed

that the AGV moves from line to arc, the steering

motor needs to be controlled to rotate over an angle.

However, the steering motor is unable to turn the

drive wheel suddenly to the desired angle. At the

same time, the steering motor itself also has

acceleration and deceleration process. So, there is a

lot of error here, and the law needs to be found out.

The error is then reduced by compensation. This

chapter deals with the two issues mentioned above.

Stage IIStage I

(a) The angle between the two paths

Stage II

Stage I

(b) The translation between the two paths

Figure 1. The trajectory analysis of AGV

(2) Trajectory planning of single wheel drive /

steering AGV

The flow chart of the AGV motion planning program

is introduced. This program is AGV's entire PC

program, and AGV can follow the planned results of

the movement. The program has the function of track

input and display. The parameters are written into the

moving program, and the attitude parameters are

feedback when the AGV is moving. First of all, users

need to design the trajectory parameters, while the

trajectory is divided into straight line trajectory and

arc trajectory. The parameters required for the line

trajectory include five aspects. They are the running

speed of the AGV, the direction of motion of the

AGV (forward or backward), the distance of the AGV

motion, the angle between the trajectory of the

The Frontiers of Society, Science and Technology

9

segment and the previous section, the direction of the

included angle (clockwise or counterclockwise). The

parameters of arc trajectory include six parameters.

They are the speed of the AGV AGV movement

direction (forward or backward), the direction of

rotation of the arc (clockwise or counterclockwise),

arc radius, arc angle, the angle of the trajectory with a

trajectory, direction angle (clockwise or counter

clockwise) a total of six parameters.

(3) Experimental result analysis

The AGV vision navigation experiment was

completed in the laboratory. Taking black belt as its

navigation road marking, its driving environment is

the cement pavement in the laboratory. In the

experiment, the running speed of AGV is set to

5m/min, and its trajectory is paved with straight line,

arc, ladder and so on. AGV can track the trajectory,

and the results are satisfactory. The parameters

obtained by a trajectory with smaller curvature and a

wavy trajectory are analyzed. Experimental results

for the first trajectory. The trajectory of this strip

consists of a straight line and a curve with smaller

curvature. Figure 2 is a result of a comparison of the

identified trajectory deviations and the measurement

deviations

0 1000 2000 3000 4000 5000 6000

-60

-40

-20

0

20

40

me

asu

rem

en

tsD

ire

ctio

n d

evia

tio

n

AGV operating distance

Deviations identified by computer

Deviations from actual

Figure 2. Comparison of identification bias and measurement deviation

As shown in the figure, the AGV begins at least in

a straight line, and the slope of the identified

trajectory is within 5S. Then, AGV moves along an

arc, and the change of the slope is that the slope

begins to grow from small to small at last. AGV

adjusts the attitude of the AGV according to the

positive and negative value that identifies the

deviation. If the identified value is negative, the

AGV moves in the positive direction, and vice

versa. By comparison, it can be seen that the

deviation of the measured AGV is basically

changed according to the detected deviation

direction. It shows that the correction effect of

AGV is satisfactory according to the result of

visual navigation.

Experimental results for the second trajectory. This

trajectory is a S curve consisting of several arcs

with large curvature. At the same time, the

response speed of AGV is improved due to the

large curvature. The trajectory deviations and

measurements identified by the computer are

shown in figure 3.

The result of the deviation comparison shows that

there is a gap between the actual running track of

AGV and the laying track. As shown in the

diagram, the starting angle of the AGV is a

negative value, that is, AGV has an angle at the

beginning of the path. As AGV runs, the angles

tend to become smaller and then change toward

positive values. It shows that AGV is in an arc, and

then the angle change is violent. It shows that the

curvature of the second section arc is larger than

that of the first section. Finally, the angle slowly

returns to zero and then moves to a stop along a

slightly curved arc. Because the response speed of

AGV is improved, the deviation correction effect

of AGV is better than that of the first experiment.

Although the response speed is increased, the error

is greater than the first experiment due to the large

curvature and rapid curvature change. But the

results of the AGV follow the guidance belt are

still satisfactory.

The Frontiers of Society, Science and Technology

10

0 1000 2000 3000 4000 5000 6000

-300

-250

-200

-150

-100

-50

0

Ru

nn

ing

dis

tan

ce

in

X d

ire

ctio

n

Running distance in Y direction

AGV guided trajectory

The actual trajectory of AGV

Figure 3. Comparison of actual trajectory and following trajectory

From the above experiments, it is shown that the

volume of AGV used in the experiment is larger.

For a larger trajectory, the tracking error is

significantly larger than the curvature of the

trajectory. Therefore, the vision navigation is

suitable for small size and fast response AGV. On

the other hand, the external conditions will have a

certain impact on the results, so the accuracy of the

image recognition needs to be improved. At the

same time, the results of this experiment are

satisfactory, which fully demonstrates the

feasibility of AGV using visual navigation. 5. CONCLUSION

The motion path of AGV trajectory is discussed in

two aspects. One is to use the path shape developed

by the user on the upper computer to plan the path of

the AGV. Another is to use visual navigation to direct

the AGV along a fixed path. In the former method, it

not only analyzes the existence of the angle between

the two tracks, but also analyzes the theoretical

solution that produces the error in turning. This not

only guarantees the movement accuracy of AGV

between different trajectories, but also provides a

good solution and reference for the future study of

using this method to formulate AGV trajectories. In

the second method, through analysis of image

enhancement, denoising and segmentation, genetic

algorithm is used to solve the trajectory recognition.

The feasibility of the theory is proved by experiment,

and the actual error is given. This provides a good

theoretical and practical guarantee for the recognition

of other images in the future.

ACKNOWLEDGMENT

The authors acknowledge the National Natural

Science Foundation of China (Grant: 51578109), the

National Natural Science Foundation of China (Grant:

51121005).

REFERENCES

[1]Vogel-Heuser, B., Rösch, S., Fischer, J., Simon, T.,

Ulewicz, S., & Folmer, J. (2016). Fault handling in

plc-based industry 4.0 automated production systems

as a basis for restart and self-configuration and its

evaluation. Journal of Software Engineering &

Applications, 9(1), 1-43.

[2]Hu, C., Jing, H., Wang, R., Yan, F., Li, C., & Chen,

N. (2015). Fault-tolerant control of fwia electric

ground vehicles with differential drive assisted

steering. IFAC PapersOnLine, 48(21), 1180-1185.

[3]Xiong, Z., Zhenhuan, Y. E., Juan, H. E., Chen, L.,

& Linghu, J. (2015). Small agricultural machinery

path intelligent tracking control based on fuzzy

immune pid. Robot, 37(2), 212-223.

[4]Pan, X., Zhou, Y., Gao, S., & Xu, M. (2015).

Virtual endoscopy automatic navigation method

based on rotation minimizing frames. Jisuanji Fuzhu

Sheji Yu Tuxingxue Xuebao/Journal of

Computer-Aided Design and Computer Graphics,

27(8), 1550-1555.

[5]Zhang, R., Li, Y., Yan, Y., Zhang, H., Wu, S., & Yu,

T., et al. (2016). Control of a wheelchair in an indoor

environment based on a brain–computer interface and

automated navigation. IEEE Transactions on Neural

Systems & Rehabilitation Engineering, 24(1),

128-139.

[6]G. Lin, X. Zou, Q. Zhang, & J. Xiong. (2017).

Visual navigation for automatic guided vehicles

based on active contour model. Nongye Jixie

Xuebao/Transactions of the Chinese Society of

Agricultural Machinery, 48(2), 20-26.

[7]Sun, Q., Teng, S., Du, Y., Kang, Z., Yin, C., &

Chen, L. (2016). Research on calibration method for

tractor automatic navigation control system. , 10(1),

129-140.

[8]Bai, X. P., Hu, J. T., Gao, L., & Zhang, T. (2016).

A sliding-mode variable-structure controller based on

exact feedback linearization for automatic navigation

system. International Journal of Agricultural &

Biological Engineering, 9(5), 158-165.

The Frontiers of Society, Science and Technology

11

[9]Cai, D. (2015). Museum navigation based on nfc

localization approach and automatic guidance system.

International Journal of Computer Applications,

120(3), 514-517.

[10]Luan, S., Qiu, T., Song, A., Wang, P., Li, J., &

Tan, H. (2017). Automatic dependent

surveillance-broadcast time delay estimation based

on extended recursive maximum correntropy

algorithm. Iet Radar Sonar & Navigation, 10(8),

1500-1507.

The Frontiers of Society, Science and Technology

12

Application of Ozone MBBR Process in

Refinery Wastewater Treatment

Henrik Rasmus Andersen1,*, Kamilla M. S. Kaarshol1, Ravi Kumar Chhetri1

1Department of Environmental Engineering, Technical University of Denmark, Miljøvej 113, 2800 Kongens

Lyngby, Denmark

*E-mail: Henrik@ndersen.dk

Abstract: Moving Bed Biofilm Reactor (MBBR) is a

kind of sewage treatment technology based on

fluidized bed. At the same time, it can also be

regarded as an efficient new reactor between active

sludge method and the biological membrane method.

The application of ozone MBBR process in refinery

wastewater treatment is mainly studied. The key

point is to design the ozone +MBBR combined

process based on MBBR process. The ozone +MBBR

process is used to analyze the treatment of

concentrated water COD discharged from the refinery

wastewater treatment plant. The experimental results

show that the average removal rate of COD is

46.0%~67.3% in the treatment of reverse osmosis

concentrated water by ozone MBBR process, and the

effluent can meet the relevant standard requirements.

Compared with the traditional process, the ozone

MBBR process is more flexible. The investment of

this process is mainly ozone generator, blower and so

on. The prices of these items are relatively

inexpensive, and these costs can be offset by the

excess investment in traditional activated sludge

processes. At the same time, ozone MBBR process

has obvious advantages in water quality, stability and

other aspects.

Keywords: Ozone MBBR; refinery Wastewater; COD

1. INTRODUCTION

In recent years, the state has increased the intensity of

environmental pollution management and pays more

attention to environmental protection issue, which

means that the country will increasingly strict sewage

discharge standards for petroleum and petrochemical

industries. With the progress of the times and the

development of science and technology, the

disadvantages of the traditional sewage treatment

methods are becoming more and more obvious, and it

has been unable to meet the current requirements of

the relevant departments for the discharge of water

quality. In order to solve this problem, the

optimization of existing wastewater treatment

processes and the development of new wastewater

treatment processes have become a pressing matter of

the moment for major petrochemical enterprises [1].

In this paper, the existing refinery wastewater

treatment methods in a chemical enterprise are

studied. When the refinery wastewater produced by

the enterprise is treated by oil separation and

biochemical treatment, the reverse osmosis process is

used to further treat the wastewater [2]. However, in

the advanced treatment, there will be a large number

of reverse osmosis concentrated shrinkage which is

not easy to degrade organic compounds. This part of

reverse osmosis concentrated water contains more

than 50mg/L of COD, which does not conform to the

standards set by relevant parts [3]. At present, the

methods to deal with concentrated water are very

scarce, and most companies have chosen reverse

osmosis advanced treatment, that is, to inject fresh

water into the water for dilution. Although this

method can reduce the content of COD in the

concentrated water to a certain extent and make it

conform to the standard, this greatly consumes the

water resources [4]. In view of this, in order to reduce

environmental risks and ensure the discharge

standards of enterprises, it is necessary to study the

treatment technology of refinery wastewater, and

build a set of devices that can effectively deal with

concentrated water [5].

2.METHODS

(1) Experimental principle

The ozone MBBR process is based on the strong

oxidation characteristics of ozone (O3). The strong

oxidation characteristics of O3 can oxidize a variety

of pollutants which are not easy to degrade, and

decompose or dissociate them into small molecules

which can be easily handled by microorganisms. The

experiment designed in this paper mainly uses the

strong oxidation of O3 to reduce the content of COD

in reverse osmosis concentrated water, so that it can

reach the acceptable level without dilution of water.

Then MBBR method is used to treat the concentrated

water treated by O3 deep oxidation process for two

times, so as to reduce the content of COD, ensuring

that the quality of the effluent meets the requirements.

It is well known that O3 is unstable in water, which is

rapidly decomposed into oxygen in an aqueous

solution with low purity. At the same time, hydroxyl

(OH) with 2.8V oxidation-reduction potential and

other small molecules with strong oxidation such as

single atom oxygen (O) are formed. Although

oxidation occurs when O3 dissolves in water, the

The Frontiers of Society, Science and Technology

13

process is relatively long. However, it has high

selectivity, and can react with ethanol and C6H5OH.

At the same time, O3 dissolves in water and causes

hydrogen chain reaction by decomposing hydrogen

radicals. The decomposition chemical reaction of O3

can be expressed by formula (1) and the chemical

reaction with water can be expressed by formula (2).

)(23 OOO +→ (1)

OHOHO 2)( 2 →+ (2)

The carrier material used in the moving bed biofilm

reactor is K3, which has the following advantages:

Biofilms grow on the inner surface of cells.

Compared with the outer surface, the inner surface is

more protected, and the biological content is

relatively fixed. There will be no shrinkage of the

surface area due to internal congestion.

Compared with the biofilm reactor filled with other

fillers, the moving bed biofilm reactor selected in this

paper has larger surface area and larger biofilm

quantity. The basic amount is 10 to 14g/L, and the

amount of VSS contained in the specific biofilm can

be as high as 95%, which can guarantee the

biological activity to some extent.

There are more species of bacteria on K3 carrier. At

the same time, the environment on the K3 carrier is

more suitable for the growth of microorganisms

which can degrade and decompose non-degradable

materials. Therefore, the effluent quality of K3

treated with MBBR is better. In addition, in view of

the biological growth mode of K3 matched with the

environmental requirements of nitrifying bacteria

growth, the number of nitrifying bacteria is relatively

heavy, which makes the nitrification efficiency more

ideal. It can reach 1 kgNH3-N/m3 D, and the

nitrification efficiency is increased by nearly 5 times

compared with activated sludge.

The ratio of K3 carrier is close to 1, so it tends to

flow when it is in the reactor, which can effectively

promote mass transfer, reduce congestion and energy

consumption.

K3 can rotate freely in the reactor, which can

effectively promote the splitting and collision times

of bubbles in water. In addition, it can prolong the

time length of the bubbles to stay in water, which

promotes the use efficiency of O2 and reduces the

energy consumption of aeration to some extent.

(2) Equipment and materials

Experimental raw water:

The pretreatment for waste water can adjust the

production of sediment by pH, and the strong

oxidation of O3 can mineralize the harmful organic

compounds in sewage, so as to remove the COD in

reverse osmosis concentrated water. At the same time,

it can transform harmful substances and COD which

cannot be decomposed by microorganisms into

organic compounds which can be degraded by

microorganisms, so as to improve their biological

characteristics. After O3 oxidation, the concentration

of COD in concentrated water will be reduced to

some extent. At this time, the use of activated carbon,

aeration, coagulation and other treatment methods to

re-care can be clearly reduce the concentration of

COD in concentrated water.

The raw water is obtained from reverse osmosis

concentrated water, which is produced by oil

separation and biochemical treatment of refinery

wastewater. The unique nature of this water is that it

contains a small amount of heavy metals and difficult

(non) biochemical degradation substances. At the

same time, it has the characteristics of low

biodegradability and toxicity to microorganisms. The

amount of water to be treated is about 0.5m3/h. The

content of COD is not more than 200mg/L, the

content of ammonia nitrogen is not more than 8mg/L,

and the suspended matter is not more than 20mg/L.

Major drugs:

The drugs used in this experiment and their purity

requirements are shown in table 1.

Table 1 Experimental drugs

Reagent Purity Manufacturer

NaOH Analytical reagent Tianjin Standard Technology Co., Ltd.

K2Cr2O7 Analytical reagent Tianjin Tianda Chemical Reagent Factory

HgSO4 Analytical reagent Jiangyan Huanqiu Reagent Factory

Ag2SO4 Analytical reagent Tianjin Standard Technology Co., Ltd.

H2SO₄ 98% Jinzhou Gucheng Chemical Reagent Factory

(NH4)2Fe(SO4)2·6H2O Analytical reagent Tianjin Standard Technology Co., Ltd.

FeSO4·7H2O Analytical reagent Tianjin North-Tianyi Chemical Reagent Factory

phenanthroline Analytical reagent Shanghai Reagent Factory

Distilled water — Self-made

H2O2 30% Shenyang Federal Reagent Factory

The equipment needed by the experiment is

assembled and assembled by the research unit and a

chemical enterprise, mainly including catalytic

oxidation tower, tail gas destruction tank and MBBR

The Frontiers of Society, Science and Technology

14

biochemical tank. Specific equipment parameters are shown in table 2.

Table 2 Parameters of the main unit device

Main unit Clearance size (m) Active area (m3) Filler content (m3)

Catalytic oxidation

tower

4.0Ф0.5 0.7232 Catalytic packing 0.45

Tail gas destruction

tank

0.35.02.1 1.8

MBBR biochemical

tank

0.32.12.1 3.8 K3 biological vector 1.5

Drain trap 2.12.1 Superficial area 1.6m2

(3) Experiment process

Experimental flow:

The specific process of this experiment is shown in

figure 1. First of all, we need to start the test and

run the water, then access to tap water after

determining that the device function is normal. The

tap water is filled into the device until it is full, and

then the sludge is connected to the MBBR pool.

Nutrients, such as P and sugar, are added for

microbial growth according to the setting time. At

the same time, aeration should be carried out

according to the actual situation. After the biofilm

is added, the raw water is injected into it, which is

about 1/5 of the total raw water. Then, the amount

of water is slowly increased and the number of

nutrients injected is reduced, ensuring that the

microorganisms in the MBBR pool can survive

normally in the pool.

Ozone generation

Flo

w m

eter

Catalytic reaction

column

Tail gas destruction

tank

MBBR biochemical

poolPrecipitate

Reverse osmosis

concentrated water

Atmosphere

Pump

Sludge return

Figure 1. Test process flow chart

After the culture of microorganisms for a period of

time, the index data of biochemical effluent can be

measured. It is confirmed that the removal effect is

ideal and the removal effect is stable. The raw water,

catalytic oxidation influent, effluent, biochemical

influent and effluent data are collected and arranged.

After the successful culture of microorganisms, the

experiment is carried out according to the

experimental process. Once the experiment can be

carried out stably, the process is adjusted according

to the experimental results. That is, under the

condition that the process conditions are not

changed, the amount of water is adjusted and the

raw water is continuously injected until the amount

of water reaches the predesigned quantity. At the

same time, the input amount of O3 should be

adjusted so as to save the input resources. O3 is

adjusted to the appropriate dose that can meet the

process requirements. Then, the DO, temperature

and other adjustable conditions are adjusted based

on the actual test results to ensure that these

conditions can guarantee the best experimental

results. After meeting the design requirements, the

impact test can be used to find the maximum

operating point. In order to supplement the

nutrients needed for microbial growth in MBBR,

the industrial grade KH2PO4 must be continuously

added after the system is stable. Every treatment of

1t wastewater requires the addition of 1.5g

KH2PO4.

The Frontiers of Society, Science and Technology

15

Parameter control:

Temperature: about 25℃.

Sludge volume: In the initial stage of the system,

nutrients should be increased to improve the sludge

concentration. After the system runs stably, the

water can be increased and the nutrient solution is

reduced until the nutrient is stopped completely.

PH value and alkalinity: The pH value of waste

water is determined by pH on-line monitor. Dosing

pump is used to automatically add Na2CO3 to

control pH value of wastewater, and pH value is set

between 6~7.

Aeration rate: The system dissolved oxygen is

maintained at 1~2mg/L during the initial start-up

and biofilm formation. When the influent load is

increased to the design load, the dissolved oxygen

is maintained at 2~5mg/L.

Ozone control: According to the experimental data,

the amount of O3 is adjusted and the dosage of O3

is reduced as much as possible to ensure the

removal rate.

Analysis method:

The water quality indexes of this experiment are

measured by national standard analysis method.

Specifically, the potassium dichromate method is

used to determine COD; the mass concentration of

NH3+-N is measured by the dichromate titration

method of sodium reagent; the pH is measured by

glass electrode method; and the content of DO is

measured by a dissolved oxygen meter.

(4) Treatment effect

After the completion of biochemical microorganism

domestication and biofilm formation on MBBR

filler, sampling analysis began in June 1, 2017, and

the analysis data were shown in table 3.

Table 3 Experimental data

Date Inlet flow

（m3/h）

Ozone

（g/h）

Oxidation influent COD

（mg/L）

Oxidation effluent

COD

（mg/L）

Biochemical effluent

COD (mg/L)

1 4.0 0 162 133

2 4.0 0 192 147

3 4.0 0 168 164

4 4.0 0 165 148

5 4.0 0 183 175

6 4.0 0 129 112

7 4.0 0 207 181

8 4.0 0 149 125

9 4.0 8.68 175 138 120

10 4.0 8.68 186 143 122

11 4.0 8.68 152 120 98.3

12 4.0 8.68 168 123 81.8

13 4.0 8.68 164 119 73.8

14 4.0 8.68 122 77.6 65.3

15 5.0 85 114 81.8 49.1

16 5.0 85 94 90 65.4

17 5.0 85 102 94 61.3

18 6.0 85 135 131 69

19 6.0 3.101 126 94 61.5

The Frontiers of Society, Science and Technology

16

20 6.0 3.101 123 77.7 85.8

21 8.0 3.101 137 80 45

22 8.0 3.101 105 55 31

23 8.0 3.101 119 75 39

24 8.0 3.101 136 73.7 44.2

25 8.0 3.101 120 63.4 38.7

26 8.0 8.90 127 69.8 41.5

27 8.0 8.90 152 80.6 48.7

28 5.0 8.60 165 80.6 37.2

29 5.0 8.60 168 85.7 43.5

30 5.0 8.60 167 79.5 39.4

31 5.0 8.60 175 82.6 43.8

32 5.0 45 127 65.4 40

33 5.0 45 126 68 42

34 5.0 45 136 78.9 36

35 5.0 45 144 74 44

36 5.0 45 152 82.3 48

37 5.0 45 146 84 44

28 5.0 45 140 78 38

29 5.0 45 134 70.2 40.9

30 5.0 45 136 76.4 40

Through the analysis for table 3, it can be found

that when the influent is normal, the influent COD

is about 141mg/L. Through the catalytic effect of

O3, the effluent is about 77. 2mg/L, and the

removal rate of catalytic oxidation is about 43.9%.

The biochemical effluent is about 41 mg/L, and the

biochemical removal rate is about 46.3%. When the

high pressure pump is opened, the influent COD is

about 170.6mg/L, and the catalytic oxidation

effluent is about 83.7mg/L. The removal rate of

catalytic oxidation is about 50.9%, the biochemical

effluent is about 41.9mg/L, and the biochemical

removal rate was about 49.9%. The effluent COD is

less than 60mg/L, which meets the national

emission standard.

The basic reason for this result is that the amount of

O3 dissolved in water increases substantially

because of the addition of O3. When the catalyst is

added, the oxidation of O3 will result in the

decomposition of many hydroxyl radicals [6]. The

higher the content of the molecule is, and, the more

organic matter it touches, which means that the

amount of organic matter that can be treated

increases with the removal rate [7]. In addition,

increasing the amount of O3 can degrade the

intermediate products accumulated in the process to

inhibit the degradation of pollutants and produce

CO2 and H2O. The degradation of these

intermediate products will be more beneficial to the

decomposition of harmful substances by

microorganisms, that is, to promote the reaction.

This can also improve the removal rate to a certain

extent [8].

After the treatment of MBBR process, using the

advantages of K3 carrier filler, the wastewater COD

can be further treated until it reaches the standard

[9]. At the same time, ozone dosage is 45 g/h, and

the power of ozone generator SKW is adjusted to

13%, that is, the power consumption of 0.65KW

tons of water is about 1 yuan. There is no blower in

power consumption equipment. Relative to the

engineering design, the operation cost is 0.11 yuan

/t.H2O. This shows that the cost of dealing with a

ton of water is relatively cheap, and the operating

The Frontiers of Society, Science and Technology

17

cost is relatively low [10].

3. CONCLUSION

In order to solve the problem that COD is not up to

standard in the concentrated wastewater discharged

from some petrochemical refinery wastewater

treatment plants, the ozone MBBR process is

adopted to treat the wastewater in depth, and the

effect of the process on the treatment of reverse

osmosis concentrated water is investigated. The

experimental results show that the process has a

significant effect on the degradation of COD. Under

normal influent conditions, the removal rate of

catalytic oxidation is about 43.9%, and the

biochemical removal rate is about 46.3%. When the

high pressure pump is opened, the removal rate of

catalytic oxidation is about 50.9%, and the

biochemical removal rate is about 49.9%. O3 can

oxidize this kind of sewage. After MBBR treatment,

the effluent quality is up to standard, and it tends to

stable. The effluent COD is less than 60mg/L,

which meets the national emission standard. At the

same time, through the calculation of the equipment

loss and investment cost in the process of

wastewater treatment, it is concluded that the

investment cost of dealing with a ton of water is

relatively cheap, and the operation cost is relatively

low.

REFERENCE

[1]Fazal, S., Zhang, B., Zhong, Z., Gao, L., & Chen,

X. (2015). Industrial wastewater treatment by using

mbr (membrane bioreactor) review study. Journal

of Environmental Protection, 06(6), 584-598.

[2]Mirbagheri, S. A., Bagheri, M., Boudaghpour, S.,

Ehteshami, M., & Bagheri, Z. (2015). Performance

evaluation and modeling of a submerged membrane

bioreactor treating combined municipal and

industrial wastewater using radial basis function

artificial neural networks. Journal of Environmental

Health Science & Engineering, 13(1), 1-15.

[3]Alkmim, A. R., Costa, P. R. D., & Ricci, B. C.

(2015). The application of filterability as a

parameter to evaluate the biological sludge quality

in an mbr treating refinery effluent. Desalination &

Water Treatment, 53(6), 1440-1449.

[4]Amaral, M. C. S., Borges, C. P., Cerqueira, A. C.,

& Torres, A. P. (2015). Treatment of refinery

effluents by pilot membrane bioreactors: pollutants

removal and fouling mechanism investigation.

Desalination & Water Treatment, 56(3), 583-597.

[5]Hoinkis, J. (2015). Membrane bioreactors for

water treatment, in: advances in membrane

technologies for water treatment: materials,

processes and applications. Journal of Water Reuse

& Desalination, 3(3), 260-267.

[6]Zaffaroni, C., Daigger, G., Nicol, P., & Lee, T. W.

(2016). Wastewater treatment challenges faced by

the petrochemical and refinery industry, and

opportunities for water reuse. Water Practice &

Technology, 11(1), 104-117.

[7]Pakravan, P., Akhbari, A., Moradi, H.,

Azandaryani, A. H., Mansouri, A. M., & Safari, M.

(2015). Process modeling and evaluation of

petroleum refinery wastewater treatment through

response surface methodology and artificial neural

network in a photocatalytic reactor using poly

ethyleneimine (pei)/titania (tio 2) multilayer film on

quartz tube. Applied Petrochemical Research, 5(1),

47-59.

[8]Rezvantalab, S., & Bahadori, F. (2015). Natural

gas refinery wastewater treatment by zeolites.

Progress in Agricultural Engineering Sciences,

11(1), 1-8.

[9]Deng, F., Qiu, S., Chen, C., Ding, X., & Ma, F.

(2015). Heterogeneous catalytic ozonation of

refinery wastewater over alumina-supported mn

and cu oxides catalyst. Ozone Science &

Engineering, 37(6), 546-555.

[10]Wang, L., Shen, X., Ni, K., & Li, X. (2015).

Treatment of oil-refinery wastewater by

heterogeneous catalytic ozonation process. Chinese

Journal of Environmental Engineering, 9(5),

2297-2302.

The Frontiers of Society, Science and Technology

18

Association Rules Redundancy Processing

Algorithm Based on Hypergraph in Data

Mining

Ankit Agrawal1,*, Alok Choudhary1

1Department of Electrical Engineering and Computer Science, Northwestern University, Evanston, IL 60201,

United States

*E-mail: ankitag@eecs.northwestern.edu

Abstract: In order to achieve the research from

individual data to data system and from passive

verification of data to active discovery, taking high

dimensional data oriented data mining technology as

the research object, an association rule redundancy

processing algorithm based on hypergraph in data

mining technology is studied according to the project

requirements. The concepts of hypergraph and

system are introduced to explore the construction of

hypergraph on 3D matrix model. In view of the

characteristics of big data, a new method of super

edge definition is adopted, which improves the ability

of dealing with problems. In the association rules

redundancy and loop detection based on directed

hypergraph, the association rules are transformed into

directed hypergraph, and the adjacency matrix is

redefined. The detection of redundancy and loop is

transformed into the processing of connected blocks

and circles in hypergraph, which provides a new idea

and method for the redundant processing of

association rules. The new method is applied to the

data processing of practical projects. The

experimental results show that the 3D matrix

mathematical model and related data mining

algorithms in this paper can find new high-quality

knowledge from high-dimensional data.

Keywords: Data mining; hypergraph; association rules;

redundant processing

1. INTRODUCTION

Energy is exhausted in consumption, which leads to

the scramble for energy and the fear of the crisis.

Data will not only be exhausted by possession, but

will continue to grow with the use and dissemination

of data. It's conceivable that data will be the DNA of

the world in the future. However, the Matthew effect

caused by data size, quality and application will lead

to obvious data resource segmentation and

competitive advantage. Therefore, big data has

become the next strategic front for all countries. This

is a war without smoke of gunpowder. With the rise

of "smart economy", the collection, mastery and

application of data will become the core

competitiveness of the state, enterprises and units [1].

Big data is a kind of data information which is

difficult to be processed or analyzed by traditional

methods or tools. With the changing times, in

addition to business data, we also record the location,

temperature, web pages and other information.

Complex data and information make up big data.

Because big data is complex, it needs enough

imagination to make up. So far, there is no uniform

definition. People’s understanding on big data is not

comprehensive and in-depth cognition. Everyone

knows only one part of big data. As for the meaning

of big data, there are several kinds of meaning at

present. Amazon big data scientist John Rosser

believes that big data refers to a huge amount of data

that exceeds the processing power of a computer. Dan

Bin, the chief product consultant of Informatica in

China, believes that big data is the sum of massive

data and complex data type [2].

We need to understand the features of big data to

better navigate this huge data. As far as we know

today, big data has 4 features: volume, variety,

velocity and value. Volume refers to the huge amount

of data, and is mainly reflected in the large amount of

data access and calculation. Data storage can jump

from the TB level to the PB level. Variety contains

structured data tables, semi-structured and

unstructured text, video, images and other

information [3]. The exchange of data is very

frequent and widespread. Velocity: On the one hand,

it means that data is constantly updated, and the

growth rate is fast. On the other hand, it means that

data has fast processing speed on data access and

transmission. Value means the value density of the

data is very low. Taking video as example, in

continuous monitoring, the useful data is only one or

two seconds [4].

Big data is not only a surge in data volume (it

gradually expands to increase the Internet data from

the initial ERP/CRM data, then includes to the related

information data such as Internet of things sensors),

but also a promotion of data complexity. Big data is a

large scale qualitative change after it is accumulated

to a certain extent. The data types of big data are rich

and varied, including the structured information such

The Frontiers of Society, Science and Technology

19

as the original database data and the unstructured

information such as text and video. Moreover, the

speed of data acquisition and processing is also

getting faster and faster [5].

Big data contains the mean of massive data and is

beyond the massive data in the content. In short, big

data is the sum of massive data and complex data

type. Big data refers to all data sets including

transaction and interactive data set. The size or

complexity degree is beyond the ability of using

common technologies to capture, manage and process

these data sets at reasonable cost and time limits [6].

Big data consists of three major technologies:

Massive transaction data: Based on the online

transaction processing (OLTP) and analysis system

from the ERP application process to the data

warehouse application process, the traditional

relational data, unstructured and semi-structured

information continue to grow. As more data and

business processes move to the public and private

sectors, the situation becomes more complex. The

internal business transaction information mainly

includes online transaction data and online analytical

data. It is a structured static historical data that is

managed and accessed through relational databases.

Through these data, we can understand what

happened in the past. Massive interactive data: This

new force is made up of social media data such as

Facebook, Twitter and LinkedIn. It includes a call

detail record (CDR), equipment and sensor

information, GPS and geolocation mapping data, the

massive image files transmitted by ManageFile

Transfer protocol, Web text, click stream data,

scientific information and e-mail. These data can tell

us what will happen in the future. Massive data

processing: A variety of light databases are used to

receive data from the client and import it into a

centralized large distributed database or distributed

storage cluster. Then, the distributed database is used

to query and classify the concentrated mass data that

stored in it. In this way, most common analytical

requirements are met. At the same time, the data

mining based on the previous query data can meet the

needs of high-level data analysis [7]. For example,

YunTable is a new generation of distributed database

based on traditional distributed database and new

NoSQL technology. It can build a hundred level

distributed cluster to manage massive data at PB

level.

According to the characteristics of big data, a new

method of super edge definition is adopted to

improve the ability of dealing with problems. Since

most mining algorithms use only support and

confidence to evaluate association rules, there is a

large amount of redundancy and loop between the

generated rules. This is particularly true when the two

threshold is set very low, which makes it difficult for

users to analyze and use these rules. To solve this

difficult problem, the concept of hypergraph is

introduced in this paper. The association rules are

represented by directed hypergraph, and the

adjacency matrix is redefined and constructed. On

this basis, the association rules loop and redundancy

detection are realized by using the properties of

directed hypergraph and graph theory, which greatly

improves the quality of association rules.

2. METHODS

(1) Brief introduction to hypergraph

At present, most of the association rule mining

algorithms use only the support and confidence to

evaluate the association rule, which cannot guarantee

that all of the mined association rules can arouse

user’s interest. It may include a large number of

redundant and meaningless association rules, which

has become one of the most important issues

affecting the efficiency of rule generation. A

mechanism for detecting redundancy and loop rules

is established, and the mechanism is used to judge the

generated Association rules. If there is redundancy or

loop rules, the rule is deleted. This method is a kind

of typical processing methods. After the generation of

association rules, the algorithm is filtered according

to the established rules of users, and the redundancy

and loop rules are deleted. Post processing method is

an important means of filtering association rules, and

has caused many experts and scholars to carry out

extensive and in-depth research. This is also one of

the important applications of this project.

Hypergraph is a subset system of finite set, which is a

generalization of graph theory and plays a very

important role in discrete mathematics. The term

"hypergraph" was first proposed by Berge in the

monograph Hypergraphs in 1966. At first, the

purpose was to generalize some classical results in

graph theory. Later, people gradually realized that

some theorems in graph theory could be generalized

in the unified form of hypergraph. Thus, the prelude

to the study of hypergraph theory has been opened up,

which has become a huge new branch of the theory.

Compared with general graphs, hypergraph research

is more complex. Some important structures and

properties in general graphs do not exist in

hypergraph, thus many similar problems of graph

theory are also more complex.

(2) Brief introduction to association rule

Association rule mining is an important branch of

data mining. It was first proposed by R.Agrawal in

1993 for the analysis of the market basket problem,

and used to analyze the shopping habits of customers.

It is used to discover interesting relationships

between items or attributes in a database. These

relationships are pre-unknown and hidden. That is to

say, it cannot be obtained by logical operations of the

database (such as the connection of tables) or

statistical methods. With the continuous collection

and storage of data, the scale of data is expanding

rapidly. Many industry people are more and more

interested in mining association rules from massive

The Frontiers of Society, Science and Technology

20

and high-dimensional databases. About the research

on association rule mining, the research department

of famous university research institutions in the

world and the major IT companies have invested a lot

of energy to carry on the exploration, and a lot of

research results are used to practical applications.

The intelligent database system laboratory of

Stanford University has developed a large number of

commercial data mining systems, such as DBMiner

mining system. It contains many advanced mining

algorithms and has the following characteristics:

Because the underlying mining details are opaque to

the user, users can use the software without in-depth

study of statistical knowledge. There are many

types of knowledge mining, including association

rules, cluster analysis and time series analysis. Due to

the adoption of advanced research result, the speed of

the product is claimed to be 20 times as much as its

competitors. In addition, the system can run on many

platforms, and is integrated with many mainstream

database systems (such as SQL-Sever and Oracle). At

the same time, the online analysis mining technology

is introduced to make the system give full play to the

advantages of data warehouse analysis.

The Quest project at IBM's Almaden labs is also a

leader in data mining research. The project includes

association rules, sequence patterns, classification

and clustering of time series. Its representative

product is DB2 Intelligent Miner of Data. This

product is applied on DB2 platform of IBM, and also

has similar products under WindowsNT.

Association rule mining is applied to massive and

high-dimensional data sets, especially in the era of

big data. It explores the real valuable information

hidden behind it. It is a long-term goal for scholars

engaged in the research of association rule mining.

Due to the particularity of massive and

high-dimensional large data, it inevitably leads to a

huge number of combinations between projects. It

takes a lot of time to search for frequent item sets.

Therefore, how to improve the efficiency of

generating frequent item sets and reduce the number

of database scanning and memory space occupied by

candidate item sets are the direction that experts and

scholars in the field of association rule mining need

to work hard. With the increasing of the dimension of

database, the search space of mining algorithm and

the possibility of blind discovery are also increased.

Therefore, we must use professional knowledge to

extract and discover tasks related data, delete data

irrelevant to task discovery, and effectively reduce

the dimension of the problem. At present, most

association rules mining algorithms use only support

and confidence to evaluate association rules. This

does not guarantee that the mining association rules

are all interested in the user. It may include a large

number of redundant and meaningless association

rules, which has become one of the most important

issues affecting the efficiency of rule generation.

The task of association rule mining in definition 1: It

refers to finding strong association rules with

minimal support and confidence provided by users in

the database D . The corresponding item set of

strong association rules →:R X Y is X Y ,

which must be a very complex item set. The

confidence of the association rule →:R X Y

derived from the item set X Y is also computed

by the support of the frequent item sets X and

X Y . Therefore, association rule mining can be

decomposed into two steps:

It is necessary to find the frequent item sets satisfying

the minimum support provided by the user in the

database.

On the basis of frequent item sets, all association

rules satisfying minimum confidence provided by

users are generated.

The basic model of association rule mining is shown

in figure 1.

Database Algorithm 1Frequent

itemsetsAlgorithm 2

Association

rules

Set the minimum support Set the minimum confidence level Evaluation

User

Step1 Step2

Figure 1. The basic model of association rule mining.

The task of Step 1 is to quickly and efficiently find all

frequent item sets in a database D . This is the

central problem of association rule mining, and is the

main index to measure the efficiency of association

The Frontiers of Society, Science and Technology

21

rule mining algorithm. At present, most of the

researches on mining association rules are aimed at

the first sub problem. It is also the main research

content of this paper: that is, to maximize the

efficiency of the algorithm.

The algorithm of Step 2 is relatively simple. It is only

an enumeration probe procedure of strong rule

produced by frequent item sets. Concrete

implementation method: For each frequent data item

set L , all its nonempty sets are found. For each

nonempty subset a , if the ratio between Support (L)

and Support (a) is greater than the minimum

confidence, then the following rules can be formed:

→ −( )a L a .

That is, a subset of any frequent data item set is used

as the premise of the rule, and the remaining part is

used as the rule conclusion.

3. REDUNDANT ALGORITHM OF

ASSOCIATION RULES BASED ON

HYPERGRAPH

(1) Redundant rule detection

At present, most association rules mining algorithms

only use support and confidence to evaluate

association rules, so that there is a large amount of

redundancy between the generated rules, especially

when the two thresholds are set very low. Rules that

are mined by association rules often contain items

that are not required by users. The rules are consistent

with the prior knowledge that the user is already

familiar with. The resulting multiple rules express the

same meaning. This kind of rules cannot provide

more useful information and any effective help for

decision makers. Therefore, it is regarded as a

redundant rule. Under the background of big data,

this situation is more serious. Redundant rules do not

bring new information. However, in many cases, the

number of redundant rules is greater than the number

of meaningful rules. The results will not only lead to

waste of mining resources, but also affect the

selection and use of association rules, thus reducing

the effectiveness of mining results. Therefore, how to

distinguish the redundant rules, and how to cut

unnecessary rules and improve the efficiency and

effectiveness of mining association rules while

ensuring the integrity of information have become

one of the hot research field. This will also be one of

the key technologies in the era of big data.

Assuming that there is an association rule

→1:R X Y and →

2:R X Y . If the rule meets

( ) ( )A B X Y and X A , them the total

number of redundant rule is − −(3 2 1)Y Y

. Y is

the number of items contained in item set Y.

The theorem proves that there are a large number of

redundant rules that can be deleted in the mining

association rules under the existing evaluation criteria.

The total number of redundant rules is analyzed

theoretically.

(2) Directed hypergraph representation of association

rules

In this paper, a method based on spanning tree is used

to remove the redundancy of association rules. This is

a new redundancy check method for association rules,

which can effectively check redundant rules:

subordinate rules and repetitive path rules. The

adjacency matrix of directed hypergraph is mainly

applied to simple graphs, where there are compound

points for directed hypergraph. Therefore, the

compound rule can only be expressed by directed

hypergraph, and the adjacency matrix must be

redefined. The adjacency matrix defined in this paper

is to take the compound point as a row or a column

just like the simple point. In this way, the information

of association rules can be accurately expressed. At

the same time, the adjacency matrix is ensured to be

the 0-1 matrix to eliminate the difficulty of the

algorithm caused by the non 0-1 matrix and the

ambiguity of the expression. After removing the

redundancy by algorithm, the adjacency matrix can

be reduced to association rules. The redefined

adjacency matrix and its directed hypergraph are

shown in figure 2.

a: Redefined adjacency matrix

b: Directed hypergraph

Figure. 2. Redefined adjacency matrix and directed

hypergraph

The Frontiers of Society, Science and Technology

22

(3) Graphical representation and processing of

redundant rules

The adjacency matrix of directed hypergraph

completely characterizes the adjacency relations

among vertices in a graph. In the directed hypergraph

based on association rules, the adjacency matrix

reflects the relation among all the association rules to

be detected.

Definition 2: The road. Assuming that there is

= ( ( ), ( ))G V G E G , and its approach is a finite

nonempty sequence =

0 1 1 2...

k kW v ev e e v

, that is, an

alternate sequence between vertices and edges. There

is ( )

ie E G

and ( )

jv V G

. ie

is associated

with −1iv

and iv

. 1 ,0i k j k

means that the point 0( , )

kv v

is regarded as the

starting point and ending point of road W .

−1 2 1, ,...,

kv v v

is called the default point of the

path W . k is called the length of W . If the

1 2, ,...,

ke e e

of road is different, then it is called

the trace. If the starting point and the ending point are

the same, the path is called the closed path (closed

track, closed circuit). Closed loop is also called

circle.

Definition 3: Connected graph. If there is a path

between u and v (two vertices) in G, then u and v are

connected. If any vertex of the graph G is connected,

then G is connected graph.

Definition 4: The tree. If there is no loop in a

connected graph, it is called a tree. According to the

definition of tree, the following properties are

obtained directly: There is at most one edge between

any two nodes in a tree. The number of edges is less

than 1 of the number of nodes in the tree. If any edge

is removed from a tree, it becomes an unconnected

graph. Adding one edge at any point in the tree will

form a loop.

From the definition of graph theory, it is found that

the processing of redundant rules in directed

hypergraph based on association rules can be

transformed into finding connected blocks in

hypergraph. Meanwhile, it also can be transformed

into spanning tree. Each edge of a directed

hypergraph represents an association rule. When the

connected graph is changed into a spanning tree, the

edges need to be deleted. This edge is the redundant

rules in association rules, and the redundant rules can

be deleted by this method.

Definition 5: Hyper tree. Assuming there is a

hypergraph ( )= = 1 2, ,{ , }

nH X ，X x x x and

= 1 2

{ , , , }m

E E E E , H is called hyper tree if it is

connected and does not contain any hyper loop.

Definition 6: Sub-hypergraph. Hypergraph H(V,E) is

defined as a new hypergraph HA=(VA,EA) by the

part hypergraph derived from the super edge subset

EA E .

In the above definition, when VA is exactly V, HA is

called the spanning sub graph of H. HA and H have

the same vertex set, but the super edge set of HA is

the super edge subset of H.

Definition 7: Connected hypergraph. The spanning

hyper tree T of connected hypergraph H= (V,E) is

defined as a spanning sub-hypergraph of H. It

satisfies the definition of hyper tree, that is, T is

connected without hyper loop.

Definition 8: Bipartite graph. The bipartite graph

corresponding to the hypergraph H= (V,E) refers to

take the vertex set V and the super edge set E of H as

the vertex set. When i iv e in H, vertex

iv is

connected with vertex ie in bipartite graph. The

bipartite graph corresponding to H is represented by

G<H>.

Graph 3 shows the relation between hypergraph and

its corresponding bipartite graph.

a: Hypergraph

b: Bipartite graph

Figure. 3. Hypergraph and its corresponding bipartite

graph

In graph theory, connectivity is of special importance,

and it can be used to map the connectivity of graphs.

Many scholars have studied it deeply since ancient

times. The directed hypergraph generated by this

project cannot be a connected graph. We need to find

connected blocks and process the spanning tree on

the basis of connected blocks. The search of

connected blocks is achieved by adjacency matrix.

From the above code, it is found that the adjacency

matrix of the spanning tree can be obtained after

The Frontiers of Society, Science and Technology

23

inputting the adjacency matrix of connected block.

The adjacency matrix of the spanning tree is obtained.

Through reduction, the deletion of association rules

redundancy is realized. Combined with the current

algorithm, we assign smaller weights to the edges

corresponding to the important association rules. The

association rules with unimportant and

non-interesting are assigned with larger weights.

Finally, the minimum spanning tree is calculated with

Prim algorithm.

In the graph, H=(V,E) and (u,v) represents the edge of

vertex u and vertex v, that is, ( , )u v E . However,

W(u,v) represents the weight of the edge. If there is a

subset of T as E, then T E belongs to a loop free

diagram. The ( )W T of

= ( , )

( ) ( , )u v T

W T W u v is the

smallest. T is called as the minimum spanning tree of

H. The minimum spanning tree is actually the

abbreviation of the minimum weight spanning tree:

1.11 11.12 10.28 10.21 10.14

Through practice, it is found that the minimal

spanning trees are different in their edge sets

sometimes. The weights are introduced to deal with

the minimum spanning tree, which plays an important

role in preserving important and interesting rules

(4) Algorithm flow

As shown in figure 4, a flow chart of the removal of

redundant association rules classification method

based on spanning tree is presented, and the steps are

as follows:

The experimental data is scanned and association

rules are generated. The directed hypergraph is used

to represent the association rules, and the adjacency

matrix is redefined.

The adjacency matrix of the pre-processing is

obtained by removing the subordinate rule algorithm.

A spanning tree without loop and connectivity is

obtained by spanning tree algorithm.

The adjacent matrix of the generated tree is reduced

to the corresponding association rules, and the final

processing result is obtained.

Experimental data

Generate association

rules

Super graph

Adjacency matrix

Preprocess adjacency

matrix

Spanning Tree

Restore rules by

adjacency matrix

Using directed

hypergraph to represent

association rules

Deletion of subordinate

rules

High quality rules

Figure. 4. Flow chart of redundancy detection for

association rules

An algorithm using redefining adjacency matrix to

remove subordinate rules is proposed. Each

association rule is defined as an edge of a directed

hypergraph, and the redefined adjacency matrix is

obtained according to the upper section. The row of

adjacency matrix represents the antecedent of

association rules, the column of adjacency matrix

represents the consequent of association rules. The

algorithm is as follows:

For each column of adjacency matrix (association

rule consequent), it is necessary to query whether its

value is 1. The intersection operations are performed

for the rows corresponding to the positions of the 1

(the antecedent of association rules). If there is

intersection, the corresponding row (the antecedent of

association rules) that contains more items for the

position 1 is set to 0. In the same way, the process is

performed for each column.

It is necessary to traverse whether the row and

column values in each item are all 0, and if so, the

row and column needs to be deleted.

Finally, the column and row are traversed from the

tail end respectively. If a column / row is all 0, then

we need to delete the column / row until the 1 column

/ row is encountered.

The output matrix is the adjacency matrix after

deleting the subordinate rule, and the composite node

in the adjacency matrix has no intersection with other

nodes. This makes it independent in subsequent

processing just like simple nodes, so it can be treated

as simple node processing.

The Frontiers of Society, Science and Technology

24

This removes all the subordinate rules in the

redundant rules. The flow chart of the algorithm is

shown in figure 5. Through this algorithm, all the

subordinate rules in the redundant rules can be

deleted, and the pre-processing adjacency matrix can

be obtained.

Association rules

Directed hypergraph

Redefine the adjacency matrix

Whether the location of

the bank is 1

Whether the position

of this column is 1

Whether there is

intersection

Whether there is

intersection

Is it the last

line?

Is it the last

column?

Corresponds to the back

of the calculation

Corresponding to the

previous operation

Set to 0 by the owner Set to 0 by the owner

Is there any

row / column all

0?

Traverse forward from the end of

the column / row

Is there a

column / row?

Preprocess adjacency

matrix

Delete the row / column

Delete traversed

columns / rows

Yes Yes

Yes

Yes

Yes Yes

Yes Yes

No

No

No

No

No

No

No

No

Figure 5. Flow chart for removing subordinate rules

4. RESULTS

In this paper, the method of removing association

rules redundancy based on spanning tree is divided

into three modules: redefined adjacency matrix

module, removing subordinate rule module and

spanning tree module. The first two modules are

The Frontiers of Society, Science and Technology

25

implemented by VB programming, and the

spanning tree module is implemented by Matlab.

The specific contents are shown in figure 6.

Redefine adjacency

matrix module

Matrix restore Spanning Tree Module

Remove the dependency

rules module

Figure 6. Module composition of classification

removing redundancy method

2 UCI data are selected, namely Balloons data set

and Shuttle-landing-control data set. The minimum

support and confidence of the Balloons dataset are

5%. There are 4 attributes in the dataset, and 18

association rules are obtained by Aprioir algorithm.

The minimum support of Shuttle-landing-control

data set is 40%, and the minimum confidence is

100%. There are 7 attributes in the dataset, and 15

association rules are obtained by Aprioir algorithm.

Taking Balloons data set as an example, the

redefined adjacency matrix is obtained according to

18 association rules (V2=SMALL is represented by

V2, others and so on).

After removing subordinate rule algorithm and

spanning tree algorithm, the spanning tree without

loop connectivity is obtained. The specific process

of spanning tree algorithm and the corresponding

directed hypergraph are shown in figure 7. In the

figure, the left half part is the spanning tree

algorithm in Matlab screenshot, and the adjacency

matrix that has completely removed redundancy is

obtained by the pre-treatment adjacency matrix (the

result of removal subordinate rule algorithm). The

right part is the change of the directed hypergraph

before and after the spanning tree algorithm.

a: Matlab simulation

b: Matlab directed hypergraph Figure.7. Matlab simulation and directed

hypergraph

Through the method introduced in this paper, two

data sets have removed the redundant rules

accurately and quickly, and the specific results are

shown in table 1.

Table 1 Properties of raw materials

Balloons Shuttle-landing-control

Total number

of association

rules

18 15

Subordination

rule

8 9

Repeat path

rule

7 2

Number of

removing

redundant

rules

15 11

Number of

residual

association

rules

3 4

5. CONCLUSION

Taking the high dimensional data oriented data

mining technology as the research object, the data

preprocessing is discussed and the problem of data

fusion is solved. A three-dimensional matrix model

is established, which lays the foundation for the

subsequent algorithm research. The method of

association rule analysis based on co evolution is

adopted, and good results are obtained in the

analysis of high-dimensional data. After

introducing the concept of hypergraph and system,

the hypergraph is established on the 3D matrix

model. The clustering analysis is realized by using

the HMETIS algorithm of undirected hypergraph.

The problem of redundancy and loop detection of

association rules is solved by directed hypergraph,

and high quality association rules are obtained.

These methods, models and algorithms have some

scalability, which lay a foundation for the analysis

of large data. In the association rules redundancy

and loop detection based on directed hypergraph,

the association rules are transformed into directed

hypergraph, and the adjacency matrix is redefined.

The Frontiers of Society, Science and Technology

26

The detection of redundancy and loop is

transformed into the processing of connected

blocks and circles in hypergraph, which provides a

new idea and method for the redundant processing

of association rules.

Some data preparation, mathematical models and

algorithmic tools proposed in this paper try to find

and analyze problems from the perspective of

system. These thinking has great significance and

the possibility of realization in big data era.

Obviously, the challenge of knowledge and tools is

also greater. We also make some useful

explorations on the observability, controllability

and global convergence, parameter optimization

and weight balance of data systems. These will be

verified and applied in the future practice.

REFERENCES

[1]Henriques, R., Antunes, C., & Madeira, S. C.

(2015). A structured view on pattern mining-based

biclustering. Pattern Recognition, 48(12),

3941-3958. DOI.org/10.1016/j.patcog.2015.06.018

[2]Shekhar, S., Jiang, Z., Ali, R. Y., Eftelioglu, E.,

Tang, X., Gunturi, V., & Zhou, X. (2015).

Spatiotemporal data mining: a computational

perspective. ISPRS International Journal of

Geo-Information, 4(4), 2306-2338.

DOI:10.3390/ijgi4042306

[3]Tang, G., Pei, J., Bailey, J., & Dong, G. (2015).

Mining multidimensional contextual outliers from

categorical relational data. Intelligent Data Analysis,

19(5), 1171-1192. DOI: 10.3233/IDA-150764

[4]Zamora, M., Baradad, M., Amado, E., Cordomí,

S., Limón, E., Ribera, J., ... & Gavaldà, R. (2015,

October). Characterizing chronic disease and

polymedication prescription patterns from

electronic health records. In Data Science and

Advanced Analytics (DSAA), 2015. 36678 2015.

IEEE International Conference on (pp. 1-9). IEEE.

[5]Xun, Y., Zhang, J., Qin, X., & Zhao, X. (2017).

FiDoop-DP: data partitioning in frequent itemset

mining on hadoop clusters. IEEE Transactions on

Parallel and Distributed Systems, 28(1), 101-114.

[6]Al-Najdi, A., Pasquier, N., & Precioso, F. (2016,

June). Frequent Closed Patterns Based Multiple

Consensus Clustering. In International Conference

on Artificial Intelligence and Soft Computing (pp.

14-26). Springer, Cham. DOI ：https://doi.org/10.1007/978-3-319-39384-1_2

[7]Hu, J., Jia, S., & Wu, K. (2015). Semantic-based

requirements content management for cloud

software. Mathematical Problems in Engineering,

2015. DOI.http://dx.doi.org/10.1155/2015/474157

The Frontiers of Society, Science and Technology

27

Research and review on the knot of Innovation

and Entrepreneurship Training Program

project in Leshan Normal University in

November 2018

Zhao Li xin1,2, 3,4,5,6*

1 Leshan Normal University, Leshan, 614000 ,China

2 Sichuan Conservatory Of Music, Chengdu, 610021, China

3Sichuan Agricultural University, Yaan,625014, China

4 Sichuan University ,Shuangliu,610207, China

5 Chengdu College of University of Electronic Science And Technology of China ,611731 , China

6 Sichuan Vocational college of Art ,Wenjiang,611131,China

*E-mail: 295887314@qq.com

Abstract: Based on the 130 topic materials in the

Innovation and entrepreneurship training of Leshan

Normal University in the 2018, this paper organizes

and analyzes the collation and depth of the functional

basis in the study of the integration of the 130

twinning materials as the research object. Objective:

After the thinking evaluation of artistic music

structure combined with 130 topic materials as the

basic research index in the deep periodic "path

analysis under the idea of music creation practice

Project" Midterm experimental report in the "X

Matrix 《金陵十三钗》" Art practice research as the

foundation of the study modelled, At the same time,

taking the practice of innovative Entrepreneurship

Training project of college students in Leshan

Normal University as the main practice project as the

main body, taking the practice of Zheng Haiyan

Music Studio in Leshan Normal University as an

example, this paper makes a series of experimental

practical functions and reviews of related materials.

Keywords: innovation; entrepreneurship; scientific

research experiments; musician scientific research;

path analysis;

1. STUDY

Generally speaking, on the analysis of the cognitive

structure of the whole concept of innovation and

entrepreneurship, on the one hand, in the formation of

the cognitive function of thinking structure, the

development of the thinking function of education

and teaching with a certain foundation is constructed,

which lays the foundation of this research. On the

other hand, in the establishment of functional basic

cognition, the feature shows the expansion of

innovative entrepreneurship in the career function. At

the same time, in the function of integrating the

structure of the integrated structure, the paper

constructs the review and integration in the study of

the structure of the horizontal function with the basic

function as the function of the career structure in the

holistic research perspective, generally speaking,

Based on the perspective of the development of

structural cognition and the comprehensive

perspective of the structure of the deep integrated

structure, taking Leshan Normal University as the

cognitive expression of the horizontal structure, this

paper studies the deep integration of the research

promoted in the development of integrated functions

as the fundamental research wholeness in the general

structure, In the promotion of the structure of the

phenotypic system in the development of the general

cognition of the structure as a basic display in the

cognitive structure of the development of the function

of cognitive integration, the formation of structural

education teaching as the subjective structure of the

hierarchical research in the basic exploration of the

construction of artistic functional development,

generally speaking, In the function of the structure of

the exhibition of Art in the past five years, the

development of the research on the integrity of the

scientific research project in the development of the

art structure in depth and the integration of the

The Frontiers of Society, Science and Technology

28

integrated structure in the general exploration is

based on the integration of the integrated structure in

the depth and the study of the subjectivity of the

study. Similarly, under the function of the integrated

structure, which is reviewed from the structural

function of biology, it promotes the functional

exploration of the research and development from the

perspective of the subjectivity of the development

cognition of the structure, and in the view of the

structure of biotechnology, the study of green

synthesis of acetylene ether in aqueous phase, the

study on the synthesis of the new method of

Quinolones derivatives, The study on breeding and

genetics of Panlong ginseng in endangered medicinal

plants and the butterfly of "using parasitic bees to

prevent and control dishes" have constructed the

integration of the thinking structure model in the

development of the cognitive model of functional

structure, Thus it shows that the study of the structure

of structural patterns formed in the field of research

as the subject of the depth of exploration and

cognitive biological research as the main body of the

description of the whole exploration of research in

the study of the development of biological patterns of

the subject, the same, Under the function of

integrating the integration sequence with the research

in the development of the integrated structure as the

main research mode, the functional continuation of

the biological machine rationality continuation of the

integration of the thinking mode in this research is

formed, on the other hand, There is also an integrated

study of the development of a model of basic

research thinking as a fundamental study of the

concept of the manifestation of biological structure

and cognitive concepts in the basic study of

integrated text in-depth thinking, Diameter structure

in the analysis of scientific research mode, the

integration of the artistic perspective of the formation

of multiple thinking dimensions is formed from the

perspective of the research on the structure of

scientific research and design, and the research on the

basis of the research from the perspective of basic

thinking is explored in depth.

2. PROFESSIONAL EVALUATION OF PROJECT

FOUNDATION AND SERVICE SPECIALTY

STRUCTURE BASED ON STRUCTURAL

CLASSIFICATION PERSPECTIVE STRUCTURE Similarly, the Project Foundation from the

perspective of evaluation-based classification

constructs more available research from the

perspective of basic research and in-depth research

and development in the study of service structure,

which is a basic assessment. On the one hand, it

continues to review the research mode in the

professional function of the service structure with the

structural project comprehensive functional cognition,

on the other hand, it studies the integrated

development of the exploration thinking structure

which is used to study the perspective in the general

research from the perspective of the integrated

classification in the structural view. This paper

constructs a comprehensive exploration of the

integrated structure of ground sequence and basic

research in the development of general cognition, In

the study of the integration structure of the general

study as the perspective of subjectivity and the study

of the development of visual threshold cognition as

the integral exploration of the integration structure,

this paper constructs the professional evaluation of

the subjectivity construction in the construction of the

general research, The formation of the general

structure of the cognitive perspective of the holistic

study and in-depth integration of the development of

functional exploration. In the research and

development of functional review from the

perspective of chemistry as the rational structure of

machine, the synthesis and properties of responsive

polymers stimulated by temperature-light-ph value In

this paper, the aesthetics of thinking structure as a

more basic cognitive development in the study of the

integrity of the responsive polymer functional group

in the research of the integration and research of the

general structure in the development of the holistic

study of the general structure of the in-depth

cognition of the TiO2 complex in the development of

the structure of the G structure symbol x function

cognition The function of the structure of the material

in the subjective material of the cognitive and

thinking of the structure of the synthesis and aesthetic

structure of the whole study in the study of the

integrity of a class of structures aesthetic perspective

of the in-depth discussion, on the one hand, This

paper fully demonstrates the study of the functional

integrity of metallic electrical factors in the function

of Rhodamine Schiff, which is induced by the study

as the subjectivity of subjectivity, and the functional

nature of the functions of the formation of the body is

studied, At the same time, in the research perspective

of the new structure, in the in-depth study, the

integration of the function of integrated structure in

the cognitive development as a perspective

exploration of the research model of the integrated

research perspective of the integration to study as the

subjectivity of the structural function in the basic

research model of the depth and development of the

overall exploration. such as "G-c3n4/tio2 composites

photodegradation Organic Pollutants Research",

"Metal ion induced rhodamine Schiff alkali

photochromic Properties Study", "Chitosan based

new adsorbent on acidic red 3 b adsorption study"

Related scientific research in the study of the factor

structure of biological structure as the study of

subjectivity in the study of the function of

exploratory structure in the integrated research of the

integration of the study of integration and functional

description in the general study of the aesthetic nature

The Frontiers of Society, Science and Technology

29

of the multi-thinking structure with the integration of

the main structure of the integrated cognition of the

exploration of the development of a comprehensive

review Integration of perspective.

3. A REVIEW OF THE PROJECT STRUCTURE OF

THE INNOVATION AND ENTREPREURSHIP

TRAINING PROGRAM IN LESHAN NORMAL

UNIVERSITY IN THE VIEW OF THE

STRUCTURE FUNCTION OF THE PATH

ANALYSIS IN THE MUSICAL CONTEXT

STRUCTURAL THINKING

Obviously, further to the test object in terms of the

structure of the experimental function, The music is

constructed in the functional group of the harmonic

structure. In the study of the multiple combinations of

the integral functional groups in the research

perspective, the integration of the integration in the

basic research perspective from the perspective of

research has been formed. The development of the

model is formed in the development of the basic

research perspective of the cognitive function of the

formation of the model of the study in the structure of

the development of the subjective innovation and

entrepreneurship of the formation of the music

context The study of structure, under the

development of holistic path analysis as the research

mode of contextual thinking of the musical structure

of subjectivity structure, constructs and forms the

integration in this research field, and forms the

deepening and development of the model of multiple

thinking structure, from the point of view of the

function of the structure of innovative

entrepreneurship, With the director of Sichuan

Conservatory of Music, Professor GeerLin, the team

went to the academic exchange of Leshan Normal

University, respectively, in the view of four operas

"Zhuo Wenjun", opera "Xue Tao", opera " Du Shi

Niang" and Opera "Yang Sheng An", and constructed

the research of basic mode as the subjectivity.

Analysis and possible sub-path model of the

continuation of the evaluation function the

exploration of the research perspective, with the

study of the processing functional structure of the

model, has formed a review of the research and

development of the integrated structure based on the

synthesis and review of the integration structure of

this structural functional factor form in the

development of the model. Moreover, from the point

of view of the function of experimental structure, the

model of the structure of the path analysis thinking is

used in the development of the cognitive structure of

the research on the influence structure of the new

media technology in the development of the

cognition of the perspective of the study on the

structural nature of the research perspective

constructed in the material cognition of the integrated

structure of the study, It has formed the integration

from the perspective of folk culture research, and the

same way, in view of the multi-dimensional thinking

research model structure of Guo Moruo culture in the

context of musical structure, constitutes the social

functional investigation of the phenomenon of

agro-industrial structure system formed as the basic

structure in the development of this research

perspective. In order to promote the innovation in the

integration research of regional structure branding

model of musical structure, The study of the holistic

structure of the exploration and development of the

general structure was formed in the exploration of the

structural function group as a review of the integrated

research and exploration of the structural functional

group of the study of the GHD structure of the study

of the structure and the exploration of the integrated

structure and the study of the integration of the

structure of the integrated model for the study of the

general structure Description of the hierarchical

structure function under the contextual thinking level

of the cognitive functional structure of the cable.

There is no doubt, obviously, the general use of "SYB

training teaching Auxiliary materials-college students

entrepreneurship story sitcom" "Wenchuan Memory:

5.12 earthquake scene report reporter video and oral

history data collection" "Leshan Yi Area Primary and

secondary school Putonghua promotion and Usage

survey and research" The functional thinking of the

research perspective of regional structure,

Constructing the holistic thinking constructed in the

perspective of the holistic study from the perspective

of general research the integration in the development

cognition of the research perspective is integrated in

the study of the development of LED as symbolic

function and the integration of the research model, as

the view of the basic structure function and the

deepening and development of the thinking structure

in the formation of this research perspective. The

exploration of sex in the study of the functional

exploration of the formation of the subjectivity of the

formation of the general structure of the study of the

research model of the development of the functional

group of the integration model, at the same time in

the "gradient type of ultra-surface properties and

application research" "high-speed serial car accident

early Warning system" "practical Electronic

measuring ruler design" Automatic Feeding Machine

Control system "solar intelligent Environmental

protection ecosystem" "The effect of crystal structure

on the luminous performance of LED fluorescent

powder" functional group, Constructs a functional

report on the core level of the functional group of this

research structure, which is based on the perspective

of the study of the functional group of clinical

structures formed in the visual angle of this research.

The exploration of the thinking structure of the

holistic development of this research perspective, as

the integration and deepening of the holistic structure

of the model of the subjectivity structure, is

The Frontiers of Society, Science and Technology

30

obviously based on the cognitive and modeling

structure of the thinking structure formed in the

development of the core structure formed in the

expansion of the horizontal structure in the field of

innovation and entrepreneurship.

In the same way, in the cognition of the development

of structural patterns, the development of research

functions based on the main perspective of study,

such as the development of the research model of

natural subjects, as the research function of the

subjectivity of the integrated structure of the factor in

the development of the integration structure of the

model of the research function, formed with the "

Investigation on the factors of adjusting the flowering

period of peony Phoenix with oil, generally speaking,

the research in this field has formed from the function

of experimental structure, which is based on the

model of studying the perspective of natural subject,

From the research of FLP structure in the integration

and development of sequence structure in the study of

abcts structure function group, this paper probes into

the synthetic functional structure of the general

Coumarin formation, taking the study as the

subjectivity of the study in the development of the

general research as the integration of the basic

structure in the formation of the " Study on synthesis

and reactivity of bridge-linked FLP "study on the

addition and reaction of green husigen ring in water

phase", in depth, in this field, it is generally

constructed to construct the development of

integrated structure in the study of structural

modeling factors of model engineering in the study of

the structure of the review as the main body of the

research report, The integration of a holistic research

perspective into the study of the core exploration of

the general research structure functional development

in the experimental exploration of "ultrasonic

synthesis of sulfonamides sodium" can be seen, based

on the continuation of the development of the

characteristics of the research perspective of the

functional factor structure of ultrasound as the main

structure. The study of the general structure as a

review function of subjectivity as the function of

subjectivity is the integration of NHR active structure

system in the path of structural mode with abcts, and

the same is done under the function of the practical

electrochemical experiment detection and analysis

structure of the structure. It also explains in depth the

research of this comprehensive structure research as

the main body of the research review model and the

development of structural functions. Obviously, in

Leshan "All the way to you-Leshan Volunteer

alliance", based on the function of service-oriented

structure of the integrity of the experiment, fully

demonstrated as the basic structure of the study in the

general exploration of integration, the construction of

this research perspective based on the integration of

the function of the deepening, policy-oriented

structure of the research perspective, " Study on

poverty alleviation project of ecological Wuji

Chicken breeding industry in Muchuan, "study on

selection and cultivation of economic tree red

Tsubaki trees", "isolation and purification of worm

kinase active bacteria in earthworm intestine" In the

study of the cultivation of the general structure and

the structure of the active functional group structure

of the biological structure in the separation and

purification structure formed in the cognition of the

poverty-alleviation function of the general chicken

structure, this research perspective is constructed in a

deep exploration of the integration of the cognitive

perspective of the separation of purification cognition

in the depth of the functional development of the

general structure , formed as a certain research

perspective as the subject of the practical structure of

the exploration of integration in the general as the

structure of college students Service project collation

and review in the general research perspective of the

holistic function of the synthetic thinking of the

structural review and in-depth exploration, the

general structure of the depth and research

perspective of the exploration, This paper forms the

possible research and in-depth review of the

possibility perspective of the new subjectivity

discoloration material in the perspective of

coexistence from the perspective of the development

and deep cognition of the holistic sequence structure

from the perspective of basic research. It has been

formed in the study of the general structure as a basic

research perspective in the study of the holistic

research perspective of the research in the perspective

of the general level of thinking of the overall

structure of the development and study of the

development of functional models and holistic

exploration. At the same time, "study on the

domestication and application of a dung straw high

efficiency methane production compound fungus",

the green Synthesis of 1,2,3-, which is the

participation of acetylene acid decarboxylase in

aqueous phase It also fully demonstrates the

integration of the integrated functions of Leshan

Normal University in the study of the function of

subjectivity structure, the function of integrated

structure from the perspective of general structure

and the in-depth development of the function of

holistic structure in the exploration of general

functional thinking review the integration of the

model function of the whole structure under the

structure function and the X function of the research

perspective in general In the in-depth study of the

structure of the integrated structure system in the

integration of the model of the development of the

overall function of the study as a fundamental

exploration of the integration and depth.

4. COMPREHENSIVE EVALUATION OF THE

INTEGRATED RESEARCH PERSPECTIVE OF

INNOVATION AND ENTREPRENEURSHIP

The Frontiers of Society, Science and Technology

31

BASED ON THE VISUAL THRESHOLD

FUNCTION STRUCTURE OF STUDIO AND

HIGHER EDUCATION TRAINING

If the study of studio model is taken as the research

of subjectivity in the research of general innovation

and entrepreneurship Big Data integration research,

this paper forms the integration in the comprehensive

structure with the evaluation as the subjectivity

research, Formation and construction of integration in

the general perspective of the training of colleges and

higher education in the view of the integrated

research model based on the integration of innovation

and entrepreneurship big Data as an integrated

exploration in the study of the research of subjectivity

research, in general, in the review report of the

development cognition based on the deconstruction

model, This paper forms a review of the function

integration in the study of the dynamic function of

research as a model in the overall structure of the

general structure. From the development of the

review research model of comprehensive structure as

a research report, For example, the function review in

the active structure explores the integration function

in the research perspective in the general oxidation

function of the soybean model in the development of

the study of the formation of the general structure of

the development of the integration of the integrated

functional land research as a review of functional

research in the general structure of the whole

exploration, In order to explore the integration

function of the integrated structure with the

integration of the basic research perspective, the

study of the integration of the integrated structures in

the experimental device, which is based on the

integration of the combination of the main study as

the research of the subjectivity, is a review, The

research model formed in the function of a certain

basic structure is formed in the exploration of the

development of functional cognition of the large data

structure of the general structure, which, as the

structural function of the basic research, In the

development of the cognitive environment structure

in the formation of this research report, the

exploration of the function of the structure of the

research expression thinking is based on the study of

the perspective as the holistic thinking level of the

general structure in the general structure of the

thinking function of the holistic thinking function of

the integration and research as a certain structure of

the development perspective of the development of

general cognition With the study as the function of

subjectivity structure, there is no doubt that the study

on the tolerance performance and regulation of

nitrifying bacteria under environmental stimulation,

"preparation of C60 and its test device in two

Chloromethane", "study on preparation and

antibacterial activity of coumarin

derivatives-nano-silica hybrid materials", is the main

evaluation project, on the one hand, Stimulated in the

present as a group of social structural functions in the

General experimental function group in the study of

the integrated structure formed by DGS in the

development of structural cognition, the integration

of this research perspective constructed in the

cognition of the general structure is integrated in the

general research function, At the same time, from the

research and development of evaluation structure as

the research perspective of subjectivity structure

under the integrated function, It has formed a review

of the integration function from the perspective of

subjectivity research. The integration and

development of integrated functional research in

general structure the study of the integration and

research perspective from the perspective of

subjectivity research as the active function of the

study of subjectivity is the integrated development of

environmental structure and the report of integrated

structure in stimulating structural functional group

performance in the perspective of relevant research

Functional research and development of the holistic

thinking assessment perspective. To see in depth,

From the perspective of the art structure, the study of

the level of thinking thought as the main body in the

development of the subject as the main body of the

research function as a certain organism in the

structural function group of the study as the

subjectivity of the study of the function of thinking

level of research in the general structure of the

perspective of the integration of the study as the

function of subjectivity thinking , the functional

transformation of the function of the integrated

structure, which is based on the perspective of

subjectivity research in the practice of Studio mode,

is formed in the study of the whole research and

exploration of the report structure as the perspective

of subjectivity research. Generally speaking, in the

study of the exploration of research functions, it has

formed the integration of structural cognition in this

field of research. From the studio of holistic research,

"innovative art education and training base" "Yi Ming

Culture and creative workshop" "Jiayi Xuan Tea

Culture Communication Studio" "Seven Color square

art therapy studio" "Ya Ya Children Handmade

headwear Studio" "Da Data Service Studio" "Jiazhou

Sunshine Youth Integration Workshop" The

functional group of the integrated structure is seen in

the art education and training base, the cultural

creative workshop and the tea Culture

Communication studio, the art therapy studio and the

decorative structure of the studio, Integration of DA

data service structure in general structure research in

the study of basic structure workshop as the

evaluation report of subjectivity structure in the

general formation of the hierarchical function of the

review structure of the research and development of

the whole, This paper forms the development of the

holistic function of the structure of the perspective of

The Frontiers of Society, Science and Technology

32

the study in the development of the functional model

of the perspective of the study as a function of a

certain economic structure in the development of the

function of increasing the value of production to

promote the development of the structure of the

overall study as an exploration of the general

structure of the development of structural integrity

and integration of the structure , it is based on the

formation of a general structure of the cognitive and

integrated structure of the deep understanding,

formed a general structure of the research perspective

in a class of experimental structure of the functional

group formed in the development of the cognitive

report formed in the research perspective of the

composition of the integration model, Has formed a

general study as a cognitive function in the

development of the integration and functional review

of the integrated research structure, which is used as

the perspective of subjectivity research. "Slyz Image

Studio" "Picking up the small Courtyard Creative

Workshop" "British Pub Cultural Experience

classroom" there is no doubt that, from the

perspective of this research function, from the

perspective of the study of the structure of the

economy as a review of the study of the subject, from

the research of the studio, This paper constructs the

cognition and development of a mental structure of

pub functional culture experience integration formed

in the functional group of this research perspective in

the study of the creative function of the holistic

structure of the exploration as the subjectivity,

forming the studio, The model takes the research

perspective in the formation of the subjectivity to

study the thinking function of the development of the

research function of the integrated structure in the

general as the formation of the research perspective

of the report in the formation of the review of

functional structures as the integration of the research

perspective of subjectivity.

5.EVALUATION AND INTEGRATION OF THE

MIDTERM EXPERIMENTAL REPORT AND THE

PHASED INTEGRATION OF THE RESEARCH

EVALUATION ON THE MUSIC CREATION

PRACTICE PROJECT UNDER THE IDEA OF

PATH ANALYSIS

From the point of view of the practice project of

music creation under the idea of path path analysis,

this paper forms a functional study, which takes the

perspective of research as the subjectivity research

structure, as a holistic study, as a review of the

cognitive development of the structure of cognitive

structure, and the development of the holistic

structure formed in the study. This paper forms the

development and integration of the whole structure of

the general structure cognitive symbol function,

which is used for the research perspective, and forms

the understanding of the holistic structure of this

research perspective in the study of the cognitive

structure of the G32 structure as a perspective

structure in the study of the functional research

perspective of the horizontal widening. In an

integrated manner, based on this study, the function

of cognitive structure formed in different stages of

experiments from the creation of works as a basic

structure in the development of exploratory research

perspective, In the study of the integrated structure

with the function of mass as the subjectivity structure

under the different hormone structure, the integrated

thinking function of the research perspective is

discussed in the development of the whole research

structure of the formation. This perspective is

constructed to explore the quality of the production

function model formed by the basic function in the

recognition of the economic cognition of the structure

of the integrated structure of the comprehensive

model in the development of the "humanization" as a

basic structure of the study of the comprehensive

function of the integrated structure formed in this

field review, The research on the integration of the

whole structure formed in this field and the research

and development in the exploration of the integrated

function of the sequence structure is the continuation

of the cognitive thinking level of the integrated

functional structure as the research perspective of the

subjectivity structure. Obviously, the project of music

creation practice under the idea of path analysis is

constructed from the perspective of this research, and

the functional embodiment of the integrated structure

of the development of the general structure and the

integration of cognition is constructed from the

perspective of the study of the main functional

formation of the holistic structure of the research with

basic continuation. There is no doubt, as far as the

understanding of deep exploration is concerned,

Taking Leshan Normal University as the research of

college students ' innovative entrepreneurship as the

subjective structure, the different hormones formed in

the seed dormancy of the treatment structure, the

analysis and research of the composition of the

research from the perspective of the study of the

function and perspective of the structure of rapeseed

lignin and fiber in the study of the structure of the

production function in the research perspective The

research perspective formed in the continuation of the

"cobweb" composition of the integrated structure in

the integrated structure of the symbolic cognition of

the structural perspective of the integration of

graduate students integrated research function group

in the experimental device structure from the

perspective of the development of subjectivity.

Obviously, as a similar structural functional group in

the evaluation effect formed in the basic study of the

perspective of structural cognition in the development

of the cognitive perspective of the integration of

biological functions in the activity of looking at your

research perspective of the structure level of

multi-dimensional integration structure, The

integration of the research and development model of

The Frontiers of Society, Science and Technology

33

the thinking level constructed in the development

from the perspective of the research of the basic

structure in this research perspective of the

construction function of hierarchical research shows

the integration of thinking and research function

in-depth and exploration of functional research

perspective of the development and evaluation of the

holistic structure of the holistic structure in the study

as a fundamental structure of the overall exploration

of the Development of cognition Cable. There is no

doubt that from the structural system of the report of

the holistic research function, "investigation on the

popularization of credit and education for college

students in Sichuan province", "a probe into the

present situation of Yi Culture inheritance in the

construction of new countryside of Yi nationality",

"Chengdu Jinli Special Snack album and

characteristic seal design conception" "Two goods"

campus WeChat Flea Market " A preliminary study

on the effect of bamboo allelopathy "investigation on

the status of traditional arranged marriages of Yi

nationality under the new era", "research on" human

yoke "in rural areas of Sichuan Province," a study on

the status quo of Gelao culture and language

protection under the great integration of nationalities

"," research on emergency investigation and

pretreatment of empty nest elderly in the region "

Marketing strategy of small Liangshan organic

agricultural products under the background

"investigation on the management of waterlogging in

Sichuan three or four-line cities" Fully constructed in

this study as the subjectivity of the structure also

functions in the general cultural function in the policy

Service category as well as in the design conception

environment of the snack structure to form the

cultural function on the one hand multidimensional

examination of the structure perspective of the

language research perspective of the level of the

study structure of the integration of functional

thinking and the study of functional structure of the

Cognitive development environment The research

perspective formed in the function integrates the

overall integration of the technical structure of the

general study. It is used to promote the deepening of

the holistic structure in the development environment

of general cognition and to explore the integration of

holistic exploration in structural cognition and the

evaluation in the cognition and research of functional

development.

6 ． TAKING THE INNOVATION AND

ENTREPRENEURSHIP OF ZHENG HAIYAN

MUSIC STUDIO IN LESHAN NORMAL

UNIVERSITY AS AN EXAMPLE,THE DEEP

PRACTICE OF BROADENING THE PRACTICE

OF MUSIC CREATION UNDER THE “PATH

ANALYSIS MUSIC THOUGHT”IS REVIEWED

Obviously, taking the innovative entrepreneurship of

Zheng Haiyan Music Studio in Leshan Normal

University as an example, the author, in helping

Zheng Haiyan carry out the innovation structure of

music entrepreneurship, has modified the practice of

using the music analysis structure as the basic path

structure in the development of advertising design

and research mode, on the one hand, To enhance the

functional value of the market structure as a basic

research perspective in the formation of the

integrated structure of the context, Similarly, the

function of stimulating college students ' innovative

entrepreneurship as the subjectivity structure is

formed in the transformation of the social function of

practical service in this research field, the

multidimensional integration of the

multi-dimensional structure of the structure is formed

in the research of the Integrated Exploration Service

of the formed structure, and the development of the

multidimensional structure is formed in the deep

economic growth. The long perspective used to break

through, obviously, is combined in the perspective of

the function of the innovative entrepreneurial

structure of college students to promote the

horizontal widening of this research perspective, it is

clear that Zheng Haiyan also graduated from Leshan

Normal University and was taught in Natalia and Olia

two outstanding Ukrainian musicians, In the

development of horizontal structure has formed the

basic function, on the other hand, based on the

characteristics of the perspective of this research

perspective assessment research, the formation of

multi-thinking hierarchy dimension of the structure of

the integration of the practice of music

entrepreneurial structure, no doubt, as a class of

structural integration of cognitive thinking level, This

paper forms the functional nature of the hierarchical

context of the basic multidimensional structure of the

study, such as the model of shared bicycle delivery in

the city, the community situation of college students

and the prevention and control of insect pathogenic

research formed in the study of the optimization

system of language training in children with autism,

which goes deep into the function of Western

structure. At the same time, it fully shows the

description of the core level of the study of the

multidimensional perspective of the holistic study of

the motivation function of Japanese majors in the

structure of enrollment. Are formed in this research

perspective of the integrated study of functional

thinking structure integration and in-depth

continuation and exploration of the mechanism of

structural functional groups in the development of the

possibility of continuity the basic nature, such as the

text analysis and the analysis of the historical

structure formed in the product development of the

new media team, the function and whole of the

hierarchical structure under the contextual thinking of

the eastern Sichuan soil text The presentation of the

model, but also shows the development of the

integrated structure in the function of the outdoor

structure of the functions of the new media team

The Frontiers of Society, Science and Technology

34

structure in the study of the functional basis of the

integrated structure in the research perspective of the

integrated thinking structure of the review and

integration, there is no doubt that the perspective of

this cultural structure has a certain integration type,

Similarly, from Sichuan Opera in the structure of the

non-legacy model, for the continuation of the level of

music thinking, "Jiajiang plank New Year's

inheritance and revitalization", "the text analysis and

historical study of the river Trumpet" "Sichuan Tujia

characteristic music instrument" gathering grass

gongs and drums "Regional Culture Research" " 《寰

宇通志》《全宋诗》 Research and inheritance of

Tibetan folk songs in Danba County, Ganzi

Prefecture "Research on the formation and teaching

mode of instrumental music associations in primary

schools"--a probe into the campus mode of the

non-legacy musical culture into the primary

school--taking Sichuan opera as an example "the

expression of network culture" culture "is integrated

from the music structure ...《全宋诗》 《寰宇通志》

7.SUMMARY

There is no doubt that, in the light of the current

situation and policy structure, from the three primary

color education resource network structure function

in the evaluation function system of the structure of

the wheat winter medicinal materials to deal with the

structural characteristics of the catalytic material in

the cognitive development of the subjective design

structure of the formation of microwave technology

treatment of catalytic materials integration,

Simultaneous formation of the integration of liquid

biochemical treatment "dr.com Assistant type"

Propagation and research the development of

integrated structure of perspective functional group

and the research on the functional system of

integrated structure as a basic research perspective of

college students in the platform structure of the

development of the model structure of the mapping of

the shape of the form of the ecological structure of

the agricultural functional group of the level of

thinking and the development of the subjective

structure of the exploration. From "college students '

self-help learning school answering platform" "The

discovery and traceability of hotspot review based on

graph", "the development and design of ecological

agriculture curriculum under the background of

research travel-taking Leshan Shawan District as an

example", "" Meng Bao Germination "early teaching

school, can be embodied in in-depth research report,

and through several stages of" path analysis "Also in

the practice of multiple thinking as the research

perspective of the main structure of the integration of

functional research review function of the

development of thinking level, formed in the normal

structure of the function of the empirical research as

the main body of research and cultivation, there is no

doubt that in this field of research perspective,

formed a multidimensional level of integration and

research perspective of the deep widening Has

formed the multi-dimensional integration of the basic

research perspective formed in the deep exploration

of the basic structure, in the function of the

development cognition of this structure, and also

constructs the integration of this research practice

report in the research function of college students '

innovative Entrepreneurship Training Project with

"Music creation practice under the path analysis

music thought" as the subjectivity structure. At the

same time, also connected with the integration of

other related entrepreneurship, and finally in-depth in

the Leshan Municipal committee secretary Peng Lin

in 2018 city at all levels of the assessment meeting,

said, Leshan at the end of 2018 sprint three goals, one,

GDP rapid and stable growth, second, the prevention

and control of African swine fever; Shawan and Big

Buddha Scenic Area traffic complete hidden danger

of sound. Obviously, in terms of GDP, Leshan

Normal University November 2018 summary of the

material materials summed up a total of 130 projects,

in terms of innovative entrepreneurial projects for

college students, it is expected to break through

GDP300 million yuan.

ACKNOWLEDGMENT

Author profile: ZhaoLixin male (199209-) Leshan

Normal University research department project

researcher（614000）; PhD,Sichuan Conservatory of

Music（610021）; Director, Research Laboratory of

Osmanthus music production and big data

communication, Sichuan Agricultural University

（625014）;PhD, Institute of Symbols and media,

Sichuan University（610207）; University researcher,

Department of Arts and Technology,Chengdu College

of University of Electronic Science And Technology

of China （ 611731 ） Choreographer,Sichuan

Vocational college of Art （611131）;

The research project is a report on the topic of music

creation practice under the idea of "path analysis" of

college students ' innovative entrepreneurship project

in the 2017;

REFERENCES

[1]Ashihara, H., Li, X., Ukaji, T. (1988). Effect of

inorganic phosphate on the biosynthesis of purine and

pyrimidine nucleotides in suspension-cultured cells

of Catharanthus roseus.Annals of botany 61,

225-232.

[2]Arbona, V., Manzi, M., Ollas, C., Gómez-Cadenas,

A. (2013). Metabolomics as a tool to investigate

abiotic stress tolerance in plants. International journal

of molecular sciences 14, 4885-4911.

[3]Baek, D., Kim, M.C., Chun, H.J., Kang, S., Park,

H.C., Shin, G., Park, J., Shen, M., Hong, H., Kim,

W.Y., Kim, D.H., Lee, S.Y., Bressan, R.A., Bohnert,

H.J., Yun, D.J.(2013). Regulation of miR399f

transcription by AtMYB2 affects phosphate

The Frontiers of Society, Science and Technology

35

starvation responses in Arabidopsis. Plant Physiol 161, 362-373