1 incremental evolution of autonomous controllers for unmanned aerial vehicles using multi-objective...

25
1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh, and Edward Grant North Carolina State University U.S. Naval Research Laboratory

Post on 20-Dec-2015

216 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

1

Incremental Evolution of Autonomous Controllers for

Unmanned Aerial Vehicles using Multi-objective Genetic

Programming

Gregory J. Barlow, Choong K. Oh, and Edward GrantNorth Carolina State University

U.S. Naval Research Laboratory

Page 2: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

2

Overview

• Problem• Unmanned Aerial Vehicle Simulation• Multi-objective Genetic Programming• Fitness Functions• Experiments and Results• Conclusions• Future Work

Page 3: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

3

Background

We have previously evolved unmanned aerial vehicle (UAV) navigation controllers able to:• Fly to a target radar based only on

sensor measurements• Circle closely around the radar• Maintain a stable and efficient flight

path throughout flight

Page 4: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

4

Problem

• We are most interested in the more difficult radar types, particularly intermittently emitting, mobile radars

• Evolving controllers directly on the most difficult radars yields very low rates of success

• We would like to create controllers able to handle all of the radar types rather than having one controller for each type

Page 5: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

5

Simulation

• To test the fitness of a controller, the UAV is simulated for 4 hours of flight time in a 100 by 100 square nmi area

• The initial starting positions of the UAV and the radar are randomly set for each simulation trial

Page 6: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

6

Sensors

• UAVs can sense the angle of arrival (AoA) and amplitude of incoming radar signals

Page 7: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

7

UAV Control

EvolvedController

AutopilotUAVFlight

Sensors

Roll angle

Page 8: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

8

Transference

These controllers should be transferable to real UAVs. To encourage this:

• Only the sidelobes of the radar were modeled

• Noise is added to the modeled radar emissions

• The angle of arrival value from the sensor is only accurate within ±10°

Page 9: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

9

Multi-objective GP

• We had four desired behaviors which often conflicted, so we used NSGA-II (Deb et al., 2002) with genetic programming to evolve controllers

• Each fitness evaluation ran 30 trials• Each run had a population size of 500• Computations were done on a Beowulf

cluster with 92 processors (2.4 GHz)

Page 10: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

10

Functions and Terminals

Turns• Hard Left, Hard Right, Shallow Left,

Shallow Right, Wings Level, No ChangeSensors• Amplitude > 0, Amplitude Slope < 0,

Amplitude Slope > 0, AoA <, AoA >Functions• IfThen, IfThenElse, And, Or, Not, <, =<,

>, >=, > 0, < 0, =, +, -, *, /

Page 11: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

11

Fitness Functions

Normalized distance• UAV’s flight to vicinity of the radar

Circling distance• Distance from UAV to radar when in-range

Level time• Time with a roll angle of zero

Turn cost• Changes in roll angle greater than 10°

Page 12: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

12

Normalized Distance

1

i

0

T

i 0

i1

fitness

idistance

distance

T

distance

distance

Tfitness

minimize tois goal The

at time distance theis

radar the to UAV thefrom

distance initial theis

steps timeofnumber total theis

1

1

2

Page 13: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

13

Circling Distance

2

i

T

iii2

fitness

in_range

N

distancein_rangeN

fitness

minimize tois goal The

otherwise 0 and range,-in is UAV when the1 is

target theof range-in steps timeofnumber theis

)(*1

1

2

Page 14: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

14

Level Time

3

i

T

iii3

fitness

level

levelin_rangefitness

maximize tois goal The

otherwise 0 and level, are wings the when1 is

*)1(1

Page 15: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

15

Turn Cost

4

i

i

T

i1iii4

fitness

iroll_angle

h_turn

roll_angleroll_angleh_turnT

fitness

minimize tois goal The

at time UAV theof angle roll theis

10 than moreby changed has angle roll theif 1 is

*1

1

Page 16: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

16

Performance of Evolution

• Multi-objective genetic programming produces a Pareto front of solutions, not a single best solution.

• To gauge the performance of evolution, fitness values for each fitness measure were selected for a minimally successful controller.

Page 17: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

17

Baseline Values

Normalized Distance ≤ 0.15• Determined empirically

Circling Distance ≤ 4• Average distance less than 2 nmi

Level Time ≥ 1000• ~50% of time (not in-range) with roll angle = 0

Turn Cost ≤ 0.05• Turn sharply less than 0.5% of the time

Page 18: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

18

Experiments

Continuously emitting, stationary radar• Simplest radar case

Intermittently emitting, stationary radar• Period of 10 minutes, duration of 5 minutes

Continuously emitting, mobile radar• States: move, setup, deployed, tear down• In deployed over an hour before moving again

Intermittently emitting, mobile radar• Most difficult radar type for evolution

Page 19: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

19

Direct Evolution

Radar TypeRuns Controllers

Total Succ. Rate Total Avg. Max.

Continuously emitting, stationary radar 50 45 90% 3,149 63 170

Continuously emitting, mobile radar 50 36 72% 2,266 45.3 206

Intermittently emitting, stationary radar 50 25 50% 1,891 37.8 156

Intermittently emitting, mobile radar 50 16 32% 569 11.38 93

Page 20: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

20

Incremental Evolution

• Environmental incremental evolution was used to improve the success rate for evolving controllers

• A population is evolved on progressively more difficult radar types

Page 21: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

21

Incremental Evolution

Radar TypeRuns Controllers

Total Succ. Rate Total Avg. Max.

Continuously emitting, stationary radar 50 45 90% 2,815 56.30 166

Continuously emitting, mobile radar 50 45 90% 2,774 55.48 179

Intermittently emitting, stationary radar 50 42 84% 2,083 41.66 143

Intermittently emitting, mobile radar 50 37 74% 1,602 32.04 143

Page 22: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

22

Comparison

Page 23: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

23

Intermittently emitting, mobile radar

Page 24: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

24

Conclusions

• Autonomous navigation controllers were evolved to fly to a radar and then circle around it while maintaining stable and efficient flight dynamics

• Using incremental evolution dramatically increased the chances of producing successful controllers

• Incremental evolution produced controllers able to handle all radar types

Page 25: 1 Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming Gregory J. Barlow, Choong K. Oh,

25

Future Work

• We have successfully tested evolved controllers on a wheeled mobile robot equipped with an acoustic array tracking a speaker

• Controllers will be tested on physical UAVs for several radar types in field tests next year

• Distributed multi-agent controllers will be evolved to deploy multiple UAVs to multiple radars