crossoverintermediate.m

遗传算法工具包

function xoverKids  = crossoverintermediate(parents,options,GenomeLength,FitnessFcn,unused,thisPopulation,ratio)
%CROSSOVERINTERMEDIATE Weighted average of the parents.
%   XOVERKIDS = CROSSOVERINTERMEDIATE(PARENTS,OPTIONS,GENOMELENGTH, ...
%   FITNESSFCN,UNUSED,THISPOPULATION,RATIO) creates the crossover children 
%   XOVERKIDS of the given population THISPOPULATION using the available 
%   PARENTS. This kind of recombination can only work for real valued
%   genomes. A point is chosen partway between the two parents. 
%
%   RATIO controls the range over which the children can occur.
%   A scalar ratio will generate children on the line between the parents.
%   A vector ratio (length = GenomeLength) can produce children anywhere in
%   the hypercube with the parents at opposite corners.
%
%   Example:
%    Create an options structure using CROSSOVERINTERMEDIATE as the crossover
%    function with default ratio = ones(1,GenomeLength)
%     options = gaoptimset('CrossoverFcn' ,@crossoverintermediate); 
%   Create an options structure using CROSSOVERINTERMEDIATE as the crossover
%    function with RATIO of .5
%     ratio = 0.5
%     options = gaoptimset('CrossoverFcn' ,{@crossoverintermediate, ratio});
%

%   Copyright 2004 The MathWorks, Inc.
%   $Revision: 1.9 $  $Date: 2004/01/16 16:51:34 $

% here's the default if nothing is passed in.
if(nargin < 7)
    ratio = ones(1,GenomeLength);
end

% How many children to produce?
nKids = length(parents)/2;

% Allocate space for the kids
xoverKids = zeros(nKids,GenomeLength);

% To move through the parents twice as fast as thekids are
% being produced, a separate index for the parents is needed
index = 1;

for i=1:nKids
    % get parents
    parent1 = thisPopulation(parents(index),:);
    index = index + 1;
    parent2 = thisPopulation(parents(index),:);
    index = index + 1;
    
    % a random number (or vector) on the interval [0,ratio]
    scale = ratio .* rand(1,length(ratio));
    
    % a child gets half from each parent
    xoverKids(i,:) = parent1 + scale .* (parent2 - parent1);
end