A jumping genes inspired multi-objective differential evolution algorithm for microwave components optimization problems

作者：Zheng, Shao Yong^*; Zhang, Sheng Xin

来源：Applied Soft Computing, 2017, 59: 276-287.

DOI：10.1016/j.asoc.2017.05.062

• Differential evolution
• Jumping genes
• Elitism leading mechanism
• Multi-objective optimization

Exploitation and exploration are equally important for multi-objective evolutionary algorithms (MOEAs) to approximate the optimal Pareto Front (PF). However, most existing multi-objective differential evolution algorithms (MODEs) focus on exploitation with proposals of elitism methods that may result in poor diversity or sticking into local optimal front. Inspired by a biological discovery, this study proposes a jumping genes based MODE algorithm, termed as JGDE with two components. The first component is the application of jumping genes operator to MODE to promote population diversity while in the second component, an elitism leading mechanism is designed to accelerate the convergence. Experimental studies show the superiority of JGDE over other competitive algorithms in both convergence and diversity. More importantly, JGDE can deal with local optimal fronts that are difficult to handle by the existing MODEs. The studies are also further verified by the solvement of a practical microwave components optimization problem and the comparison between different optimization algorithms.

• 出版日期2017-10
• 单位中山大学