An adaptive multi-objective differential evolution algorithm based on evolutionary process information

摘要

In this paper, an adaptive multi-objective differential evolution algorithm based on evolutionary process information, named AMODE-EPI, is proposed to improve the searching performance. In AMODE-EPI, the process information is used to describe the schedule of evolution. Meanwhile, the parameters, including scaling factor, crossover rate and population size, are adjusted dynamically based on EPI. Then, this proposed AMODE-EPI can balance the local search and the global exploration abilities. Finally, the performance of AMODE-EPI is validated and compared with other state-of-the-art multi-objective evolutionary algorithms on a number of benchmark problems. The experimental results show that the AMODE-EPI has better convergence and diversity than the other algorithms.