多算法多种群协同优化差分进化算法

Multi-algorithm and Multi-population Co-optimization Differential Evolution Algorithm

研究算法融合、多种群协同进化是应用群智能算法求解复杂工程应用问题的一个方法。设计了一个基于差分进化算法的多算法多种群协同优化算法,注重多算法的选择与组合。设计了一种自适应参数差分进化算法,选择了3种各具特点的差分进化算法变体与其互补,基于4种算法的特点设计了相应的多种群协同进化策略。对算法进行了仿真设计,仿真结果表明该算法通过使4种不同特点的算法互补能得到较优结果,并获得精度、可靠性与适用性的提升,弥补工程应用中算法选择的困难。

Algorithm fusion or co-evolutionary with multi populations are the solutions for complex engineering application. A multi-algorithm and multi-population collaborative optimization algorithm is proposed by differential evolution （DE） algorithm, which pays emphasis on algorithm selection and combination. The algorithm designs a parameter-adaptive DE algorithm and selects three different DE algorithm variants which is complementary for each other and provides a multi-population co-optimization scheme according to four algorithms characters. Stimulation results show that the proposed algorithm could make four different algorithms remedy for each other, gets a better result, and raises the precision, reliability and suitability, which reduces algorithm selection difficulty in engineering application.