Beruflich Dokumente
Kultur Dokumente
Content
Overview
Swarm Particle Optimization (PSO)
Example
Swarm Intelligence
Overview
Swarm Intelligence
Swarm Intelligence
http://www.scs.carleton.ca/~arpwhite/courses/95590Y/notes/SI%20Lecture%203.pdf
Swarm Intelligence
Swarm Intelligence
Swarm Intelligence
Swarm Intelligence
Swarm Intelligence
Particle Swarm Optimization (PSO)
Basic Concept
The Inventors
Russell Eberhart
James Kennedy
electrical engineer
social-psychologist
pbest k
w2 d
vk
v
v
sk
gbest k
w1 c1 rand ()
w2 c2 rand ()
sk 1
d
k 1
pbest k
gbest
gbest
pbest k
vk
vk
v
sk
sk 1
d
k
11
pbest k
w1 c1 rand ()
w2 c2 rand ()
gbest k
gbest k
pbest k
k 1
i
k 1
i
s v
k
i
v v
k
i
k
i
v w1d
k
pbest k
w2 d
vk
vk
v
sk
gbest k
sk 1
d
k
11
pbest k
w1 c1 rand ()
w2 c2 rand ()
gbest k
gbest k
pbest k
PSO Algorithm
**
Swarm Intelligence
Particle Swarm Optimization (PSO)
Examples
Schwefel's Function
n
f ( x) ( xi ) sin( xi )
i 1
Simulation Initialization
Summary
Iterations
gBest
416.245599
515.748796
850
"sample.dat"
800
750
700
10
759.404006
15
793.732019
650
600
550
20
834.813763
100
837.911535
500
450
400
5000
837.965771
Optimun
837.9658
16
64
256
1024
4096
Exercises
Facts
Background
Typical Applications
Scheduling problems
ACO Concept
ACO System
Pheromone Intensity
ij (t )
ij (0) C , i, j
(
t
)
.
ij
ij
k
pij (t )
il (t ) .il
lJ ik
, 0
ij 1/ dij
visibility
(
t
)
.
ij
ij
k
pij (t )
il (t ) .il
lJ ik
, 0
ij 1/ dij
= 0 : a greedy approach
= 0 : a rapid selection of tours that may not be optimal.
Pheromone Update
k
k
Q
/
L
(
t
)
(
i
,
j
)
T
(t )
k
ij (t )
otherwise
0
Q: a constant
Tk(t): the tour of ant k at time t
Lk(t): the tour length for ant k at time t
m
ij (t ) ijk (t )
k 1
ij (t ) (1 ) ij (t ) ij (t )
Demonstration