MATLAB: GA optimization problem, terminate before stall limit

gagenetic algorithmMATLABoptimizationOptimization Toolbox

The solution can be seen to be improving but it terminates before the generation or stall limit. Can find a better solution manually by trial and error.

Best Answer

You can set larger values for various stopping criteria, such as Generations or StallGenLimit. Or you could do yourself a favor and try using patternsearch instead of ga. patternsearch is usually faster and more reliable, and is easier to tune. If you need to start from a wide variety of initial points, and you have finite bounds on all components, try
 x0 = lb + rand(size(lb)).*(ub - lb);
Alan Weiss
MATLAB mathematical toolbox documentation
Related Question