Custom data types and constraints with genetic algorithms

Question

nc 2020-11-25

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/662548-custom-data-types-and-constraints-with-genetic-algorithms

编辑： Shlok 2024-8-21

Hello everybody,

I'm trying to implement a genetic algorithm that works with custom data types in MATLAB. Each chromosome of the population is made of two parts: a string and a vector of integers.

The string part is the binary representation of a number and I need to perform admissibility checks over it after crossover and mutation operations in order to avoid that the number represented could exceed a certain value.

I've followed this example for working with custom data types but I still have a question: how do I pass the max value as parameter to crossover and mutation functions? And, more generally, is there a way to pass any extra parameter to these functions?

For now I'm using global variables, but this feels a very bad idea and I'd like to find out which is the right way to do what I need.

Thanks in advance!

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Shlok 2024-8-21

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/662548-custom-data-types-and-constraints-with-genetic-algorithms#answer_1502644

编辑：Shlok 2024-8-21

在 MATLAB Online 中打开

Hi nc,

I understand that you want to pass additional parameters, to functions such as “crossover” and “mutation” in your genetic algorithm. While using global variables might work, there's a more elegant and safer approach: using anonymous functions.

Anonymous functions allow you to wrap your existing functions and pass extra parameters seamlessly. Let’s assume you want to pass “maxValue” as an additional parameter. Here’s how you can modify your code to include “maxValue” for both the “crossover” and “mutation” functions:

maxValue = 50; % Define your maxValue here 
options = optimoptions(options, ... 
    'CrossoverFcn', @(parents, options, NVARS, FitnessFcn, thisScore, thisPopulation) ... 
    crossover_permutation(parents, options, NVARS, FitnessFcn, thisScore, thisPopulation, maxValue), ... % passing maxValue here
    
    'MutationFcn', @(parents, options, NVARS, FitnessFcn, state, thisScore, thisPopulation, mutationRate) ... 
    mutate_permutation(parents, options, NVARS, FitnessFcn, state, thisScore, thisPopulation, mutationRate, maxValue), ... % passing maxValue here
    
    % Other options as needed 
);

In this modification:

The “CrossoverFcn” and “MutationFcn” are now anonymous functions. They call your existing “crossover_permutation” and “mutate_permutation” functions, respectively, while also passing “maxValue”.
Ensure your “crossover_permutation” and “mutate_permutation” functions are updated to accept this new parameter.

For example:

function xoverKids  = crossover_permutation(parents,options,NVARS,FitnessFcn,thisScore,thisPopulation,maxValue)
    % Your custom crossover function
    % Use maxValue here 
    % ... 
end