Dynamic Neighborhood Adjustment Strategy for Multi-Objective Evolutionary Algorithm Based on Decomposition

Multi-objective evolutionary algorithm based on decomposition (MOEA/D) has achieved great success in the field of multi-objective optimization. It decomposes a optimization problem into number scalar sub-problems. Each sub-problem is optimized by using information from its neighboring Therefore, neighborhood size each plays an important role MOEA/D. Different sizes are tested this paper. Experimental results demonstrate that larger helps achieve better convergence and diversity with more CPU time vice versa. MOEA/D uses constant during whole process, it difficult to balance convergence, running time. paper propose The adjusts dynamically different generations sub-problems reduce while similar or than other state-of-the-art algorithms. Compared original MOEA/D, experimental show adjusting good way significantly maintaining diversity. Furthermore, proposed compared five algorithms outperforms others efficiency performs similarly