【摘要】 为了进一步提升差分进化算法的优化性能，结合粒子群(PSO)算法的进化机制，提出一种混合多重随机变异粒子差分进化算法(DE-PSO)。所提算法不仅使用粒子群差分变异策略和DE/current-to-rand/1变异策略提高算法的全局探索能力和增加种群的多样性，而且还使用一种简单有效的扰动策略帮助粒子跳出局部最优值因而避免早熟收敛。选取5个基准测试函数的进行仿真实验。结果表明，DE-PSO算法在搜索精度和收敛速度都优于DE算法和PSO算法，同时，将所提算法应用到微电网能量调度效果也最优。更多还原

【Abstract】 In order to further improve the optimization performance of DE algorithm, combined with the evolution mechanism of particle swarm optimization(PSO), a hybrid multiple random mutation particle differential evolution algorithm(DE-PSO) was proposed.The proposed algorithm not only used PSO differential mutation strategy and DE/current-to-rand/1 mutation strategy to improve the global exploration ability and increase the diversity of the population, but also used a simple and effective perturbation strategy to help particles jump out of the local optimal value and avoid premature convergence.The simulation results show that DE-PSO algorithm is better than DE algorithm and PSO algorithm in search accuracy and convergence speed.At the same time, the energy scheduling effect of the proposed algorithm is also optimal.更多还原