Year: 2023
Type: book-chapter
Abstract: In MMOP there are some algorithms that are difficult to find more equivalent Pareto optimal solutions,and unable to balance the performance of solutions in decision space and objective space. In addition, the solution is too densely distributed on the boundary. A multimodal multi-objective differential evolution algorithm based on hybrid strategy (MMODE_HS) is proposed to solve above these problems. First, an effective binary tournament selection mechanism based on special crowding distance is designed to select individuals which performance better diversity in decision space and objective space and balance the performance of the solution in two spaces; Second, The reverse vector mutation strategy is introduced to reduce the quantity of boundary point, which can help the algorithm improve the distribution of Pareto subsets. MMODE_HS is compared with other multimodal multi-objective evolutionary algorithms, the results demonstrate that the proposed algorithm can search for more complete Pareto subsets on nine test problems, and effectively balance convergence and diversity.
DOI: 10.3233/faia231033
Authors: Renyun Liu, Yunhan Yang, Yifei Yao, Fanhua Yu
Institutions: Changchun Normal University, Beihua University
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4389731561