You are on page 1of 5

Proceedings of the 8th

World Congress on Intelligent Control and Automation


July 6-9 2010, Jinan, China

Design and Implementation on Cultural Algorithm in Unit


Commitment Optimization
Mu-Xun Zhou 1, Zheng-Chu Wang2, Yun-Xia Luo3, Wan-Liang Wang4
1)

School of Physics & Electronic Engineering, Taizhou University, Taizhou, Zhejiang, China (zhoumx940@126.com)
2)

3)

School of Mechanical Engineer, Taizhou University, Taizhou, Zhejiang, China (wangzc@tzc.edu.cn)

Electrical Engineering Department, Zhejiang College of Water Conservancy and Hydropower, Hangzhou, Zhejiang, China
(luoyx@zjwchc.com)

4)

School of Computer Science & software, Zhejiang University of Technology, Hangzhou Zhejiang, China (wwl@zjut.edu.cn)

AbstractThe paper develops research of feasibility and validity of Culture AlgorithmCA in unit commitment optimization. The
mathematic model of the optimization problem is established. Realization method of CA is presented, including design of population
space and culture space, design of acceptance function and influence function, and design of evolution strategies for two spaces. Solving
steps of CA is realized, and Optimization calculation program based on MATLAB language is compiled. The results of example shows
that CAs perfect advantage is faster speed and better convergence performance, which is viable and efficient for unit commitment
optimization.
KeywordsCultural Algorithm, Unit Commitment Optimization, Population Space, Culture Space, Acceptance Function, Influence
Function

1 2 3 4
1)

3)
4)

MATLAB

1.

60374056

Y505360

[1- 3]

978-1-4244-6712-9/10/$26.00 2010 IEEE

3320


[4-5]

Genetic AlgorithmGA
[6]

3.1

Hybrid Particle Swarm Optimization, HPSO

Reynolds 1994 [10]

[7-8]Culture AlgorithmCA

GA

2.

[9]

min Q = min q j u j

Belief Space

j =1

1 K

P = p ju j

(Update)

j =1

u j p min
p j u j p j min
j

u j p j max p j u j p

max
j

Population Space

1 NUM
(Generate)

14 Q m 3 / s
M
j = 1,2," , M p j j

1 CA

kW
q j j m 3 / s
u j

j u j = 1
u j = 0 P
p min
p max
j
j
j
p j min p j max j

3.2

3.2.1

14

p min

P M

2 M
u p
i i = (u i1 , u i 2 ," , u iN , pi1 , pi 2 ," , piN )

3321

P 5
i = (1,0,1,1,0; p1 ,0, p 2 , p3 ,0) 25

( 0 < < 1 )

0134

ait > s it

p a s it

i =1

p a

pi = P
popu NUM

0.6-1, p a 0-0.5

t
i

num t i popu

s it

t 0 =300

popu it = ( popsit , fit it )


t
i

t
i

t =10 =0.98-1.0
t
i

pops popu i fit

K S ,

3.2.3
A

S s it t i

Generate

S s1t ," , sit ,", s Kt sit = ( popsit , ait )

t t

t < t 1 t > t 1

a it t

a it =

t
si

t
s min
t
s min

F F
t
s max

finess = K Q k1 p1 k 2 p 2

Fstmax F

= F
t

+ (1 ) F

t
savg

p1 = max{( pi pi min ), max(0, pimin pi )} 9

i =1

p 2 = max{( pi pimax ), max(0, pi max pi )} 10

a it t i

i =1

Fsit t i Fstmin

K f k 1k 2

t Fstmax t

34

t
t
Fsavg

[0,1]

K
NUM K K
K L < K < K U

3.2.2

Update
a1 a 2

i popu it fit it t 1
1

Fstmin

1
1, fit it > Fstmin
Accept ( ) =
0, otherwise

K
s kt +1 = a1 * s it + a 2 * s tj

K s tj+1 = 0.4 * sit

0
t

si ai

3322

s kt +1 = 0.6 * sit

K
sit +1 = sit

K sit =

3.3
2

q/m3.s-1

p/mW

1#

2#

3#

10

16.0

13.0

10.7

20

25.3

23.4

22.3

30

34.7

33.3

34.1

40

44.1

44.7

46.5

50

54.2

56.6

60.7

2
50mW
1# 54.2m3/s
1#

3# 54.12m3/s
0.08m3/s 50mW
7000m3 4m3/kW.h
1750

1
CA
CA GA
GA
CA
80mW
3 4
3 4 CA
GA GA
20 CA CA
GA
2 CA

CA 30
CA

[9]
3 50mW
1
100m 0-150mW

MATLAB
100 7

20 620
0.8 0.02 20
3 CA

3323

pj/mW

P/mW

1#

2#

3#

1#

2#

3#

m3.s-1

30

28.99

1.01

33.28

50

47.65

2.35

54.12

70

50.00

20.00

76.50

80

47.57

29.96

2.47

87.41

90

49.97

31.50

8.53

98.16

100

50.00

36.26

13.74

109.50

115

50.00

40.00

25.00

127.05

130

49.97

49.40

30.63

144.91

145

50.00

49.90

45.10

164.12

[2]

[3]

[4]
4 GA

[5]

[6]

[7]

CA

MATLAB
CA

GA

Systems ,vol.2,no.2,pp.339-350,1987.
Tong S K, Shahidepour S M, Ouyang Z, A Heuristic
Short-Term Unit Commitment, Transactions on Power
System, vol.6, no.3, pp.1210-1216, 1991.
Wang S J, Shahidehpour S M, Kirschen D S.
Short-Term Generation Scheduling with Transmission
and Environmental Constraint s Using an Augmented
Lagrangian Relaxation, IEEE Trans on Power Systems ,
vol.10,no.3,pp.1294-301,1995.
,. .
,vol.22, no.1, pp. 27-29, 2006.
,.
.,vol.27, no.5, pp.5-9, 2008.
,.
.,vol.33, no.5, pp.47-50, 2007.
.(Memetic Algorithm) .
,vol.26, no.11, pp.1-5, 2007.

[8] P. Moscato, On evolution, search, optimization, genetic


algorithms and martial arts: towards memetic
algorithms,
Tech.
Rep.
Caltech
Concurrent
Computation Program. Pasadena, California, USA,
pp.1-67, 1989.
[9] ..
,vol.18, no.1, pp.64-66, 1998.
[10] Robert R G, An introduction to cultural algorithms, In:
Proceedings of the 3rd Annual Conference Evolution
Programming. pp. 131-136, 1994.

[1]

Q/

Snyder W L, Powell J H D, Dynamic Programming


approach to Unit Commitment, IEEE Trans on Power

3324

You might also like