针对标准差分进化算法缺乏全局与局部搜索的动态调整,进化后期由于种群多样性降低,算法极易陷入局部最优的问题,采用控制参数的余弦函数调整策略,并提出综合分布参数概念,根据种群个体进化过程的特点及算法的阶段特性,自适应调整算法的控制参数及进化策略.通过对工程实例的模拟计算,其结果表明了改进的差分进化算法在解决水库优化调度问题上的有效性.将改进差分进化算法的计算结果与POA及标准差分进化算法进行比较分析,充分体现了改进差分进化算法的优点.
Standard differential evolution(DE) algorithm is lack of dynamic adjustment between global search and local search,the diversity of species is decreased in the later stage of evolution.The algorithm can easily fall into local optimum.In light of these problems,launch the adjusting strategies of the cosine function to control parameters,and proposed the parameters concept of integrated distribution according to the feature of individual evolution and phase characteristics of the algorithm,adaptively adjust to control parameter and evolutionary strategies.Through simulating an actual project,the results show the effectiveness of the improved differential evolution(IDE) in solving the reservoir optimal scheduling.Comparisons of the results of IDE,progressive optimality algorithm(POA) and DE fully reflect the advantage of the IDE.