粒子群算法Matlab工具箱！

madio

PSOt, particle swarm optimization toolbox for matlab.
4 L; {2 y% V4 D
May be distributed freely as long as none of the files are
modified.
, O2 m+ {! }5 f0 }4 z- p; m
Send suggestions to bkbirge@yahoo.com

Updates will be posted periodically at the Mathworks User
Contributed Files website (www.mathworks.com) under the
) u4 k" q! f/ ^* a! WOptimization category.

# R1 I/ u  }6 X3 _- Z- GTo install:
Extract into any directory you want but make sure the matlab
2 F  G: |1 x0 ~+ Z$ _  h# _path points to that directory and the subdirectories
'hiddenutils' and 'testfunctions'.
" M9 V) c1 o" F  o
Enjoy! - Brian Birge
: I% u) P  l9 x" W2 A
-------------------------------------------------------------
-------------------------------------------------------------

4 Y' C4 }1 I3 M! H/ I# [% DINFO
Quick start: just type ... out = pso_Trelea_vectorized('f6',2)
and watch it work!
- ]" Q5 G" h. t5 Z" e4 ^; l1 x! b% r
This is a PSO toolbox implementing Common, Clerc 1", and
8 u7 w* B# ]* t! |0 Y* J+ [! G& _Trelea types along with an alpha version of tracking changing
, `( D+ Z' z) o  a1 L  aenvironments. It can search for min, max, or 'distance' of
user developed cost function. Very easy to use and hack with
( F' h$ z: p( b8 J7 v" R+ b, xreasonably good documentation (type help for any function and
2 x9 {7 k+ b5 j4 K  u% M0 e: yit should tell you what you need) and will take advantage of
vectorized cost functions. It uses similar syntax to Matlab's
optimization toolbox. Includes a suite of static and dynamic
test functions. It also includes a dedicated PSO based neural
& m6 d1 p1 O2 l; S5 J9 enetwork trainer for use with Mathwork's neural network toolbox.

Run 'DemoPSOBehavior' to explore the various functions, options,
2 @3 I5 U5 `; p* |and visualizations.

) ^$ J0 I3 s% d  C0 E3 g# TRun 'demoPSOnet' to see a neural net trained with PSO
(requires neural net toolbox).

4 i% D2 w# Y% G) s
This toolbox is in constant development and I welcome
/ s( e, t# i- w" r; ^; R, Hsuggestions. The main program 'pso_Trelea_vectorized.m' lists
various papers you can look at in the comments.
/ D6 ?: \5 M# t- y1 E/ B. K
Usage ideas: to find a global min/max, to optimize training of
neural nets, error topology change tracking, teaching PSO,
investigate Emergence, tune control systems/filters, paradigm
1 `' @0 W0 `, h' s0 b3 c0 Ifor multi-agent interaction, etc.
1 s( ~0 D; j8 \; s0 i( S6 R
( }0 a3 @6 b) m  T7 e-------------------------------------------------------------
+ J  n6 [$ R- s1 \- C-------------------------------------------------------------

& A+ J' L$ d# r' C& k6 ^. u  c
Files included:


" H  ^7 y1 N8 G# U** in main directory:
1 l. {! i* z; v+ q1 O
0) ReadMe.txt - this file, duh
. e; L/ g! M% E1 U1) A Particle Swarm Optimization (PSO) Primer.pdf  -  powerpoint converted to pdf presentation explaining the very basics of PSO
, P) x5 }, y" o7 d. R; @/ p$ X! p2) DemoPSOBehavior.m - demo script, useful to see how the pso main function is called
/ ]' d+ o7 R! Z+ P. m3) goplotpso4demo.m - plotting routine called by the demo script, useful to see how custom plotting can be developed though this routine slows down the PSO a lot
4) goplotpso.m - default plotting routine used by pso algorithm
5) pso_Trelea_vectorized.m - main PSO algorithm function, implements Common, Trelea 1&2, Clerc 1", and an alpha version of tracking environmental changes.


, G/ K$ r. u7 ]$ n2 L4 S
** in 'hiddenutils'

1) forcerow, forcecol.m - utils to force a vector to be a row or column, superseded by Matlab 7 functions I believe but I think they are still called in the main algo
2) normmat.m - takes a matrix and reformats the data to fit between a new range, very flexible
: r- W. ]7 M5 X- F2 o3) linear_dyn, spiral_dyn.m - helpers for the dynamic test functions listed in the 'testfunctions' directory

; P# ?% Z1 h5 Z: v# q5 G+ x

** in 'testfunctions'

A bunch of useful functions (mostly 2D) for testing. See help for each one for specifics. Here's a list of the names:

Static test functions, minima don't change w.r.t. time/iteration:
8 n, |/ v0 Y, s' b/ q" R 1) Ackley
/ U; W! D% A) S8 Q- u 2) Alpine
3) DeJong_f2
8 e+ g( S7 \2 }, L, ]( z 4) DeJong_f3
( R$ H5 J6 u- n( U/ c. ^! t: y8 ] 5) DeJong_f4
6) Foxhole
9 o7 R& k# F9 C! I7 n 7) Griewank
: E. g8 _' \: m; e2 y 8) NDparabola
9) Rastrigin
10) Rosenbrock
/ Z. X7 ], N6 \2 W& J  g11) Schaffer f6
12) Schaffer f6 modified (5 f6 functions translated from each other)
13) Tripod

$ }2 }6 ^$ v3 h. C2 b0 rDynamic test functions, minima/environment evolves over time (NOT iteration, though easily modifed to do so):
8 Z* @. Q, S5 ^! P14) f6_bubbles_dyn
* q3 P1 ]) D' A& ]1 H- J15) f6_linear_dyn
16) f6_spiral_dyn



7 J7 z8 X# s; n** in 'nnet' (all these require Matlab's Neural Net toolbox)

) u, s) a% K/ m- N( {) l; r 1) demoPSOnet - standalone demo to show neural net training
2) trainpso   - the neural net toolbox plugin, set net.trainFcn to this
3) pso_neteval - wrapper used by trainpso to call the main PSO optimizer, this is the cost function that PSO will optimize
4) goplotpso4net - default graphing plugin for trainpso, shows net architecture, relative weight indications, error, and PSO details on run
" }  o" ^. J3 k& T/ W
6 y- ~; q; B3 p! W  x

枫叶下的你

发表回复嗯嗯嗯ヽ(○^^)
0 V0 r0 O, l3 K/ q. O

1150865889

回复一个赚体力

2268733244

太好用了，博主牛牛牛

932624145

感觉很不错的

a602466161

标标标标
) H+ g; v% d4 S$ U' J2 ?; j1 j

57475207

发表回看看感觉挺好用复