Simulation

Experimentation

Dynamic Systems and Control

Alleyne Research GroupDepartment of Mechanical Science and Engineering, College of Engineering, University of Illinois at Urbana-Champaignhttp://arg.mechse.illinois.edu/

Applied Controls Research

ARG

Theory Simulation

Experimentation

What we do & where we are going

Developing Great Research and Great People

Careers in Industry

Understand the system

dynamics

Develop control theory

Validate via simulation and

experimentation

Produce novel control

techniques

Start with an important problem

Theory

0 100 200 300 400 5000

2

4

6

8

10

Time [sec]

Superh

eat

[C]

Superheat Response to 400 RPM Step Decrease

FF PID 1000 RPMic

PID 1000 RPMic

Right: Using an advanced PID control strategy which incorporates feedforward control, we achieve better regulation of evaporator superheat than with a conventional PID controller. Better regulation of superheat is critical to the efficiency and safety of A/C&R systems.

0 500 1000 1500 2000 2500 3000 3500 4000 4500 50003

4

5

6

7

8

Te

mp

era

ture

(C

)

Set Point Temp

Inlet Temp

0 500 1000 1500 2000 2500 3000 3500 4000 4500 50003

4

5

6

7

8

Te

mp

era

ture

(C

)

Time (sec)

Above: An experimental air-conditioning and refrigeration (A/C&R) test stand is used to validate models and test our control and optimization algorithms.Right: By optimizing the upper and lower bounds of a hysteretic on-off controller to minimize energy consumption, savings on the order of 5% are achieved. The upper figure shows the on-off cycling pattern with fixed hysteresis bounds, and the lower figure shows the optimized on-off cycling pattern.

A/C & R Optimization and Control

Vapor Compression Systems

Electrohydrodynamic Jet Printing

Roll-to-Roll E-Jet System

E-jet printing is a technique that uses electric fields to deliver an ink to a substrate for high resolution (<1µm) patterning applications. The advantages of the E-jet process are the rapid fabrication of fine structures with feature sizes down to 100nm combined with a range of material inks that can be used. Advancements in sensing and control of E-jet printing are being pursued because of its potential to fabricate high-resolution patterned devices in flexible electronics as well as biological applications such as protein arrays and biosensors. Ongoing work to develop a roll-to-roll system for E-jet printing will increase throughput, decrease costs, and enable printing on flexible substrates.

Desktop E-Jet Printer with Multisyringe

Roll-to-Roll Manufacturing5 mm

Open Quadrants Closed Quadrants

Open QuadrantsMulti-Feature Size

Micro Additive Manufacturing

ARG is home to several energy system andmanufacturing experimental platforms. Newexperimental systems are developed frequentlyand students have access to many platformswithin the lab and the school of engineering.

Control Hierarchy

SimulatedAircraft

ExperimentalHardware

MissionData

0 1000 2000 3000 4000 5000 6000 7000 80000

10,000

20,000

30,000

40,000

Mission Time [sec.]

Alt

itu

de [

ft.]

Landing& Ground

DescentClimb

Cruise

Ground Takeoff

Electrical Control System

Thermal Control System

Information Flow

Vehicle Level

System Level

Coordination Level

Component Level

Physical Level

System

60 sec

10 sec

1 sec

0.1 sec

0.001 sec

continuous

1C

2,1C 2,2C

3,1C 3,2C 3,3C

4,1C 4,2C 4,3C 4,4C 4,5C4,6C

5,4C5,3C5,2C5,1C Inputs/Outputs

Thermal/Electrical Power Systems

MATLAB/Simulink is used extensively to developand test controllers prior to implementation onhardware.

ARG has developed many toolsets over the years,several of which are actively used in industry.

Control Hierarchy

SimulatedAircraft

ExperimentalHardware

MissionData

0 1000 2000 3000 4000 5000 6000 7000 80000

10,000

20,000

30,000

40,000

Mission Time [sec.]

Alt

itu

de [

ft.]

Landing& Ground

DescentClimb

Cruise

Ground Takeoff

Electrical Control System

Thermal Control System

Information Flow

Vehicle Level

System Level

Coordination Level

Component Level

Physical Level

System

60 sec

10 sec

1 sec

0.1 sec

0.001 sec

continuous

1C

2,1C 2,2C

3,1C 3,2C 3,3C

4,1C 4,2C 4,3C 4,4C 4,5C4,6C

5,4C5,3C5,2C5,1C Inputs/Outputs

Aircraft PowerFlow Toolset Thermosys Toolset

Hydraulic Power Systems

k k+1 k+2 ….. k+pSample

time

Prediction Horizon

FUTUREPAST

Reference TrajectoryMeasured OutputPredicted OutputPast Control InputPredicted Control Input

Model Predictive ControlModel Predictive Control

Extremum Seeking

Control Signal

SampleTime

Iteration Iteration

SampleTime

0 N-1 1 N0

1

j

j+1

0

1

j

j+1

Error

ILC Algorithm

uj ej

uj+1

Systemuj+1

ej+1

Continual process improvement through process repetition

Iterative Learning ControlIterative Learning Control

Theory is central to the development of novel control techniques,and the Alleyne Research Group (ARG) utilizes various forms ofcontrol theory to advance control in many different fields.

The newest experimental system seeks toemulate the interaction between thermal andelectrical systems of an aircraft.

Electrohydrodynamic Jet Printing

Roll-to-Roll E-Jet System

E-jet printing is a technique that uses electric fields to deliver an ink to a substrate for high resolution (<1µm) patterning applications. The advantages of the E-jet process are the rapid fabrication of fine structures with feature sizes down to 100nm combined with a range of material inks that can be used. Advancements in sensing and control of E-jet printing are being pursued because of its potential to fabricate high-resolution patterned devices in flexible electronics as well as biological applications such as protein arrays and biosensors. Ongoing work to develop a roll-to-roll system for E-jet printing will increase throughput, decrease costs, and enable printing on flexible substrates.

Desktop E-Jet Printer with Multisyringe

Precision Motion Control

The Alleyne Research Group (ARG) works at the intersection of theoretical control,dynamic modeling and simulation, and hardware-based experimental validation byapplying a wide variety of control techniques to solve important societal problems.

How does it work? Why does it work?

Will it always work?

How good is our model?

How does the system behave? How can we make the system behave better?

How does it work on a real system?