Intelligent optimization algorithm for global convergence of non-convex functions based on improved fuzzy algorithm

The traditional global convergence optimization algorithm is prone to premature convergence and slowconvergence in the face of complex non-convex function. To this end, a new intelligent optimization algorithm based on improved fuzzy algorithm for global convergence of non-convex function is proposed. The general model of the optimal problem is designed and the general model of non-convex function is established. The genetic algorithm is used to optimize the non-convex function, and the global convergence of the current non-convex function is analyzed. It is found that the global convergence of non-convex function is actually based on the optimization of crossover probability and mutation probability to decide the convergence of genetic algorithm, so as to improve the global convergence. A fuzzy controller is designed, which determines the input and output variables and their membership functions, establishes fuzzy rules and anti-fuzzing process to control the crossover rate. The fuzzy control of mutation rate is similar to the crossover rate, but the difference is that the new fuzzy control rule is needed. The experimental results show that the proposed algorithm can effectively optimize the global convergence of non-convex function.