Research on Mathematical Statistics Model Algorithm under Improved Algorithm

The current evolutionary algorithms for solving multi-objective optimization mainly consider how to deal with the optimization of multiple conflicting objectives, and rarely consider the problem of processing constraints. Therefore, we propose a differential evolution algorithm for elite retention and non-dominated solution set migration operations in the evolution process to ensure the diversity of Pareto optimal solutions for multi-objective optimization problems. The performance of the basic differential evolution algorithm is improved through group chaotic initialization, adaptive crossover and mutation operations. The simulation results of three classic test functions show that the algorithm in this paper is better than the non-dominated sorting genetic algorithm in terms of uniformity, approximation and convergence speed, and the convergence speed is also better than the other two improved evolutionary algorithms.