Implementation And Improvement Of Multivariate Selected-subset Mean Feature Algorithm

In reality,most engineering problems can be modeled as a multivariate multimodal function problem,because of the complexity of the multimodal function optimization problems,the most optimization algorithm can find out the global optimal,at present the Multivariate selected-Subset Mean Feature algorithm collection average smY(xj)characteristics,characteristics of iterative segmentation more extreme value interval unipolar value of extreme value range is obtained,and use the width selectionsmYw(xj)identification of extremum.This algorithm can avoid the shortcoming of the traditional multi-peak function optimization algorithm,which can only find the global optimum.But it also has the disadvantage of low computational efficiency.In this paper,Multivariate selected-Subset Mean Feature algorithm(MSMF algorithm)is applied to realize multi-peak function optimization,and the disadvantages of MSMF algorithm are optimized and tested.The specific research content is as follows:1.This paper summarizes the current research status of multi-peak and extremum optimization,summarizes the theoretical basis of Multivariate Selected-Subset Mean Feature Algorithm,and gives the reasons why it is better than other similar optimazation algorithms.2.The optimization problem of binary multi-peak function is solved by using Multivariate Selected-Subset Mean Feature Algorithm,and the program realization process and function modules are analyzed and summarized.Several extreme functions with typical characteristics were tested and the test results were summarized.3.The problems in the test were analyzed,and then the segmentation parameters and search algorithm in Multivariate Selected-Subset Mean Feature Algorithm were improved according to the causes of the problems.The parallel computing theory is used to improve the efficiency of the algorithm4.In order to verify the effectiveness of optimization,several extreme functions with typical characteristics were tested by comparing the representative niche PSO Algorithm,and the test results were summarized.Facts have proved that the above improvement measures are effective.