Multi-objective Optimization Algorithm Based on Uniform Design and Differential Evolution

Jinrong He,Dongjian He,Aiqin Shi,Guoliang He

DOI: https://doi.org/10.1007/978-981-15-5577-0_14

2020-01-01

Abstract:The multi-objective optimization problem is an important research direction in the field of optimization. Because the traditional mathematical programming method often cannot achieve the optimal global solution, the researchers introduced the heuristic method into the multi-objective optimization problem. The heuristic method is a method of searching based on empirical rules, which can get the optimal solution or solution set of problems in the limited search space. In this paper, we proposed a multi-objective evolutionary algorithm based on uniform design and differential evolution, which use the uniform design table to construct the weight vector and utilize the crossover in differential evolution and mutation process to replace the simulated binary intersection and the simulated polynomial variation. Compared with the classical algorithm, the experimental results show that the improved algorithm is superior to the original algorithm.