如何使遗传算法能够收敛（Matlab）

Question

I am running genetic algorithm using global optimization toolbox in Matlab. I want to allow the algorithm to run until the difference between the best value is <=10^{-15}. I tried to use ftol and ctol for this purpose but the algorithm terminates at 541 iterations. Any suggestions will be appreciated!

An example of my objective function is below:

function y=objfun(x)
  t = [3,227,342,351,353,444,556,571,759,836];
  n= length(t);
  sumt = sum(t);
    y = -(- x(1)*(1-exp(-x(2)*t(n)))  + n*log(x(2)));
end

and the GA code is

options.TolFun=1e-15;
%options.TolCon=1e-15;
format longG
[x,fval] = ga(@objfun,2,[],[],[],[],[10 0.001],[20 0.1])

Answer 1

Here are a full list of stop criteria you can modify:

% default options
options = optimoptions('ga');

% stop criteria
options = optimoptions(options,'MaxGenerations', MaxGenerationsVal); % number of generations
options = optimoptions(options,'MaxTime', MaxTimeVal); % time limit
options = optimoptions(options,'FitnessLimit', FitnessLimitVal); % fitness limit
options = optimoptions(options,'MaxStallGenerations', MaxStallGenerationsVal); % stall generation
options = optimoptions(options,'MaxStallTime', MaxStallTimeVal); % stall time limit
options = optimoptions(options,'FunctionTolerance', FunctionToleranceVal); % function tolerance
options = optimoptions(options,'ConstraintTolerance', ConstraintToleranceVal); % constraint tolerance

[x,fval] = ga(@objfun,2,[],[],[],[],[10 0.001],[20 0.1],[],[],options);

I haven't tested the code. Let me if this helps :)