基于最优高斯随机游走和个体筛选策略的差分进化算法

针对差分进化算法开发能力较差的问题, 提出一种具有快速收敛的新型差分进化算法. 首先, 利用最优高斯随机游走策略提高算法的开发能力; 然后, 采用基于个体优化性能的简化交叉变异策略实现种群的进化操作以加强其局部搜索能力; 最后, 通过个体筛选策略进一步提高算法的探索能力以避免陷入局部最优. 12 个标准测试函 数和两种带约束的工程优化问题的实验结果表明, 所提出的算法在收敛速度、算法可靠性及收敛精度方面均优于EPSDE、SaDE、JADE、BSA、CoBiDE、GSA和ABC等算法, 在加强算法探索能力的同时能够有效地提高算法的开发能力.

Differential evolution based on optimal Gaussian random walk and individual selection strategies

To solve the problems of poor performance in exploitation of the differential evolution(DE) algorithm, a new DE algorithm with fast convergence rate is proposed. Firstly, the optimal Gaussian random walk strategy is used to improve the exploitation ability of the algorithm. Then, the simplified crossover and mutation strategy based on the individuals’ optimization performance is employed to realize the evolution operation so as to improve the performance of local search. Finally, the individual selection strategy is proposed to avoid local optimum and enhance the exploration performance. Experimental results of 12 unconstrained benchmark functions and two constrained engineering design optimization problems show that the proposed algorithm is superior to the algorithm of EPSDE, SaDE, JADE, BSA, CoBiDE, GSA and ABC in terms of convergence rate, stability and convergence accuracy. The proposed algorithm can effectively enhance the exploration performance and improve the exploitation ability.