You are on page 1of 3

44

2008 44 34

Computer Engineering and Applications


CHENG Le

210097

223003
Department of Mathematics and Computer ScienceNanjing Normal University
Nanjing 210097
China
Department of Computer Science and Engineering
Huaian College of Information Technology
Huai
anJiangsu 223003
China
E-mailCL211282@163.com
CHENG Le.New bionic algorithmcockroach swarm optimization.Computer Engineering and Applications20084434
44-46.
AbstractBy feeding behavior simulation of cockroachesa new Cockroach Swarm OptimizationCSO has been presented in this
s
paper.The algorithm fully utilizes social equality and swarm intelligence of cockroaches.Food distribution again and returning nest
strategy enhance global and local search capability of CSO.The simulation to traveling salesman problem proves that CSO is
superior to Discrete Particle Swarm OptimizationDPSO.
Key wordsCockroach Swarm OptimizationCSO
food distribution again
Traveling Saleman ProblemTSP
Discrete Particle Swarm
OptimizationDPSO

Cockroach Swarm Optimization
CSO

TSP
PSO
Discrete Particle Swarm Optimization

DOI
10.3778/j.issn.1002-8331.2008.34.012
834-0044-03

Ant Colony Optimization


ACO[1-2]
[3-5]
Particle Swarm OptimizationPSO
Artificial Fish-school
Algorithm[6-7]

TP242.6

[8]
C

TSP

ACOPSO Algorithm
NP

Cockroach C

TSP CSO
TSP S=s1
s2

sn
TSP

n TSP CSO S

C
J.Halloy C

n n
S C

6 TSP C

the National Natural Science Foundation of China under Grant No.60673102the Natural Science
Foundation of Jiangsu Province of China under Grant No.BK2006218

1979-

2008-06-23

2008-09-04







A421
653 B3
216
54

21

CSO Ci Fpi

Step

Step C TSP Stepx


y

26

TSP x y
C A 4
2
1
6
5
3
C Step1
6
B 1

22

45

2008 44 34

Road

FpC=Road
FgC=Road

1
2

FgRandStep=Fp

1 C Fp Road2 C Fg

Road C Step Road =


Step1+Step2++Stepm C A 421653

Road3

D 123456 2 Road=DA=

Step1
3Step3
6Step4
6

n TSP
CSO

1 Ci=1m
Fpi=1m

Ci=
i
i
Fpi=1m
Fpi=1m

i
i

Step13

Fg
2 FORi=1
im
i++

Step
 36

Step
6
 1

23

11 C Fpi
FpiC1
FpiC2
FpiC3
FpiCm

2 2 1 Ci=1m

Fpi 4

Step4
6



CSO

Fg

2
C


Food

1
Fp

Fp2

FgFood global
Fg C
FpFood partial
Fp C
Fg m C
Fpi=1m

RandStep C Fp 3

Ci=1m

Fg
i
Fg=Fp1=Fp2==Fpm

Ci Fpi Fg Step

Fpi

C1

Fpm

C2

Ci

Cm

C Fpi

3 Ci=1m

i
3 Fpi=1m
Fg
i
4 FORi=1
im
i++

RandStep

RandStep

RandStep

. . . . . . .
.........

Fp1  Fp2

Fpm

12 C Fg

FgC1
FgC2
FgC3FgCm
2 4 1 Ci=1m

Fpi 5
Fg

24

TSP C
Step
n TSP n
Step TSP TSP
n Step PSOACO CSO
C Fp Fg C
TSP TSP Fp Fg

C Fg Fg RandStep
C Fp C

25

C C C

Fp1

Fp2

Fpi

. . . .. . . .

C2

C1

Fpm

.
.......

Ci

Cm

C Fg

3 Ci=1m

i
5 Fpi=1m
Fg
i
6 Fg=Fp1=Fp2==Fpn
3
Fpi=1m
3
i
7
2

46

Computer Engineering and Applications

2008 44 34

CSO

Fg

668.58

659.65

621.75

585.52 550.70

525.89

508.31

494.27

482.21

478.87 473.71

471.94

471.07

458.63

457.79

457.63 454.62

425.65

587.33

566.03

489.68 467.21

455.82

Fg
678.00 636.03 612.11 569.94

553.29

475.12 427.88 424.69 423.91

423.74

645.03 584.96 564.10 536.03

513.60

495.82

487.60 483.81 479.99 476.47

473.62

472.76

460.31 459.48 459.06 429.78

425.48

601.48 499.14 499.14 476.94

441.13

435.36

441.13

435.36

641.11 600.43 563.88 531.46

498.24

488.19

482.18 473.70 467.69 462.13

458.45

457.32

453.33 452.21 449.71 438.05

437.19

424.74

538.60

423.74
630.18
2

448.55

428.42

423.74

679.95

582.19

533.29

424.74

423.91

423.74

447.09 447.09

432.12

657.12

622.12

595.03

552.06 536.91

525.50

515.26

506.41

501.69

470.92 466.86

462.13

458.39

438.85

429.74

424.90

551.77

551.77

526.47

515.70 507.50

425.94

425.10

423.91

423.91 423.74

430.68 425.99

423.74

601.48 499.14 499.14 476.94


9

490.48

430.68 425.99

10

423.91 423.91

423.74

423.74

2 CSO PSO

Algorithm

Run1

Run2

10

Run10

Ave

Err%

CSO

200

423.74

PSO

1 000

432.66

428.00

423.74 423.74 425.27 423.74 425.48 423.74 423.74 423.74

424.49

0.18

423.74

425.27 425.48 423.95 423.74 426.31 423.74 423.74 423.91

425.25

0.36

Run3

Run4

Run5

Run7

Run8

Run9

TSP CSO

TSP Oliver30
PCPentium

Run6

4 -2.93 GHz CPU256 M RAM

Win XP OSVC++ 6.0[9] ACO Oliver30

ACO
PSO

CSO CSO

423.74423.74 Oliver30
C Fp

Fg 1 C 200
10 Fg

[1] Colorni A
Dorigo MManiezzo Vet al.Distributed optimization by
Life.AmsterdamElsevier Publisling
1991
134-142.

1 CSO
2 CSO DPSO
10
4
Ave-423.74
Err=
100%
423.74

ant colonies[C]//Proc of the 1st European Conference on Artificial


[2] Dogigo M.Optimization
learning and natural algorithms[D].Italy
Po
litecnico diMilano1992.
[3] Hendlass T.Preserving diversity in particle swarm optimization[M].
Lecture Notes in Computer Science
2003
2718
4104-4108.

CSO C 200
PSO
1 000DPSO [10]
CSO
DPSO
DPSO

[4] Kennedy JEberhart R C.A discrete binary version of the particle


swarm algorithm[J].Proceedings of IEEE Conferenceon Systems
1997
54104-4108.
[5] Eberhart R
Kennedy J.A new optimizer using particles swarm the
ory[C]//Roc Sixth Intemational Symposium on Micro Machine and
Human Science
Nagoya
Japan.Piscataway
IEEE Service Center
1995
39-43.
[6] .

5
CSO
1 C
C C 2
3

[J].20022211
32-38.
[7] .[J].

200381
126.
[8] Halloy J.Individual discrimination capability and collective deci
sion-making[J].Journal of Theoretical Biology2006239313-323.

CSO C
C

[9] .[J].

C
C

[10] .[J].

[8] CSO

19991036
1240-1245.
20044
477-480.

You might also like