In real life and engineering practice,multiple targets to be optimized often appear at the same time.A large number of such problems are solved by evolutionary algorithm.Therefore,the study of multi-objective evolutionary algorithm has important theoretical and practical significance,and has become a research hotspot in recent years.However,when solving problems with small or discontinuous feasible regions,most algorithms cannot converge to the Pareto frontier due to the obstruction of the infeasible region.At the same time,the fixed mutation parameter makes the excellent solution and the inferior solution have the same mutation probability,which cannot meet the requirements of preserving the excellent solution and improving the inferior solution as much as possible to improve the convergence and diversity of the algorithm in the evolutionary process.In order to solve the above problems,this paper makes the following innovations and work:1.In order to help the algorithm cross the infeasible region and converge to a smaller feasible region,a multi-objective evolutionary algorithm LCMO based on the improved Logistic operator is proposed.The algorithm uses a co-evolution framework.When improving the optimizer algorithm of the framework,the improved Logistic chaotic mapping is introduced into the environment selection.The improved Logistic operator improves the uneven distribution of the original operator iteration points and can change the solution individual The distribution in the space helps the population to cross the infeasible region to improve the convergence of the algorithm.At the same time,the algorithm also uses the secondary sorting method to improve the non-dominated sorting operation to further refine the level of the non-dominated solution.Comparing LCMO with 4 advanced MOEAs experiments,the results show that the algorithm LCMO has outstanding performance on the problem with a small feasible region,and has good diversity and convergence.2.In order to make the individual mutation probability change adaptively with evolution,a multi-objective evolutionary algorithm AKn EA based on adaptive mutation operator is proposed.Based on the Kn EA algorithm,the adaptive mutation probability is introduced into the algorithm.In the environment selection stage,the mutation probability of the individual in the next iteration is calculated by comparing the fitness value of the individual and the maximum fitness value of the current population.This parameter can adaptively adjust the mutation operation according to the individual of the solution and the degree of evolution of the population.By comparing the AKn EA algorithm with four advanced MOEAs,the AKn EA algorithm achieves better results on most problems.Through the comparative experiment between AKn EA and Kn EA,it can be seen that the improved algorithm has better diversity and convergence.In general,the introduction of adaptive mutation operator improves the performance of the algorithm.3.The original location model P median model is improved,and the improved algorithm is used to solve the model and show the results. |