细菌觅食算法与粒子群优化算法相结合

细菌觅食算法与粒子群优化算法相结合的程序，可实现适应度函数的最值计算。

%% Tunning of PID controller using Bacterial Foraging Orientec by Particle swarm optimization 
%
%
%My work has been accpepted in GECCO 2008 as Graduat Student workshop. I
%have used this techique in PID tunning and i got better result thatn BG
%and PSO
%
% Author: Wael Mansour （wael192@yahoo.com）
%
% MSc Student Electrical Enginering Dept 
% Faculty of Engineering Cairo University Egypt





%%
%Initialization
clear all   
clc
p=2;                         % dimension of search space 
s=10;                        % The number of bacteria 
Nc=10;                       % Number of chemotactic steps 
Ns=4;                        % Limits the length of a swim 
Nre=4;                       % The number of reproduction steps 
Ned=2;                       % The number of elimination-dispersal events 
Sr=s/2;                      % The number of bacteria reproductions （splits） per generation 
Ped=0.25;                    % The probabilty that each bacteria will be eliminated/dispersed 
c（:1）=0.5*ones（s1）;       % the run length
for i = 1:s 
    Delta（:i）=（2*round（rand（p1））-1）.*rand（p1）;  
end 
for m=1:s                    % the initital posistions 
    P（1:111）= 50*rand（s1）‘;
    P（2:111）= 0.2*rand（s1）‘;
   %P（3:111）= .2*rand（s1）‘;
end                                                                  
c1 =1.2;                                        %  PSO parameter C1 1.2
c2 = 0.5;                                       %  PSO parameter C2 .5
R1 = rand（ps）;                                 %  PSO parameter
R2 = rand（ps）;                                 %  PSO parameter
Plocal_best_position=0*ones（psNc）;            %  PSO                       
Pglobal_best_position=0*ones（psNc）;           %  PSO
velocity = .3*randn（ps）;                       %  PSO                        
current_position=0*ones（psNc）;                %  PSO

%%
%Main loop 
    

%Elimination and dispersal loop 
for ell=1:Ned
    

%Reprodution loop


    for K=1:Nre    

%  swim/tumble（chemotaxis）loop   

        for j=1:Nc
            
            for i=1:s        
                J（ijKell）=tracklsq（P（:ijKell））;         

% Tumble
                    
                Jlast=J（ijKell）;
                Jlocal（ij）=Jlast; 	             	
                P（:ij+1Kell）=P（:ijKell）+c（iK）*Delta（:i）; % This adds a unit vector in the random direction            
 
% Swim （for bacteria that seem to be headed in the right direction）     
                
                J（ij+1Kell）=tracklsq（P（:ij+1Kell））;  
                m=0;         % Initialize counter for swim length 
                    while m                          m=m+1;
                          if J（ij+1Kell）                             Jlast=J（ij+1Kell）;    
                             P（:ij+1Kell）=P（:ij+1Kell）+c（iK）*Delta（:i） ;  
                             J（ij+1

 属性            大小     日期    时间   名称
----------- ---------  ---------- -----  ----
     文件        6320  2008-06-14 01:04  BG-PSO-matlab\BG_PSO-algorithm.m
     文件         723  2008-06-11 20:38  BG-PSO-matlab\tracklsq1.m
     目录           0  2008-06-14 01:07  BG-PSO-matlab\