Research On Multi-objective Evolutionary Algorithm Based On Reference Boint And Mutation Inertia Weight

This paper first introduces the concept of multi-objective optimization,summarizes the development process of the algorithm,and briefly introduces the evaluation index of the algorithm.Then we introduce the work done mainly in multi-objective evolutionary algorithm,that is,multi-objective evolutionary algorithm based on reference point and mutation inertia weight.The multi-objective evolutionary algorithm based on catastrophic inertia weight mainly introduces the mutation inertia weight mechanism to overcome the local optimum of the algorithm.It makes the inertia weight increase suddenly or rapidly attenuates,uses the minimum distance and the congestion distance strategy to maintain the foreign archiving set,and selects the global optimal position through the roulette strategy,and proves its effectiveness in the experiment.In the experiment,the advantages and disadvantages of the inertia weight mechanism of fixed value and linear change are compared,and the comparison between the convergence and the diversity is made using the diagram.The traditional evolutionary algorithm will meet the problem of difficulty in solving the high dimension and multiobjective problem.The second algorithms in this paper mainly study this problem.In the iterative process,a series of reference points with convergence and diversity are used to guide the evolution according to the current population,and a series of reference points are generated in the target space according to the solution scheme of the external file,and the final particle is chosen as the solution by using the relative position between the non dominant solution and the reference point.At the end of this paper,the two algorithms of this paper are used to calculate an actual use case,the economic dispatch problem of the power system,to illustrate how to apply the evolutionary algorithm to the reality.