부산대학교 도서관

Multi objective optimization evolutionary algorithms (MOEAs) play a crucial role in addressing multi-objective optimization problems (MOPs) in the field of artificial intelligence. However, MOEAs often struggle to simultaneously maintain good convergence and diversity in Many-objective optimization problems (MaOPs). To address this problem, this paper proposes a Many-objective optimization evolutionary algorithm based on dynamic region partitioning approach (MaOEA-DRP). After non-dominated sorting, the solutions are initially uniformly partitioned into regions based on Euclidean distance. Within each region, at most one solution closest to the ideal point is selected. Subsequently, dynamic region partitioning is applied to the remaining individuals. In each iteration, a solution closest to the ideal point is chosen within the obtained region until the population size is satisfied. Experimental results indicate that MaOEA-DRP outperforms other algorithms significantly in terms of performance on the test problems.