Why there is Difference in the Solutions of global/ local optimum value obtained by genetic algorithm and global search using fmincon?

I have solved an optimization problem using fmincon solver. I doubt if its a global minima, so i used genetic algorithm (GA). With GA, the solver resulted into local minima. While, when I used global search technique with fmincon, the solution obtained was same as those obtained while i used fmincon alone. Why global search and fmincon giving same values and why GA stucks in the local minima? Which should be taken as global minima?

Answer 1

[Matt J]

Perhaps you were wrong to assume that fmincon alone was failing to find the global minimum. You can never be certain that a global optimum will be found by any method, but it would make sense to pick the solution with the lowest objective function value.