多种群差分进化算法及在柔性车间作业调度中的应用

A multi-population differential evolution algorithm and its application in flexible job shop scheduling

针对柔性车间作业调度问题,在深入分析差分进化算法的基础上,提出了一种多种群差分进化算法.该算法基于DE/rand/2/bin变异方式全局搜索能力强,鲁棒性好,和DE/best/2/bin变异方式局部搜索能力强,收敛速度快;包含若干个普通种群和一个精英种群,普通种群采用DE/rand/2/bin变异方式,精英种群采用DE/best/2/bin变异方式,普通种群和精英种群及普通种群之间在适当的时候进行优秀个体迁移,以实现全局搜索能力和收敛速度之间的平衡,并从现实生产实际出发,建立了双目标柔性车间作业调度数学模型.最后,将该算法应用于一个调度算例,仿真结果表明,该算法可行有效.

In view of the problem in flexible job shop scheduling problem, a multi-population differential evolution algorithm was presented, based on the thorough analysis of differential evolution （DE） algorithm. Based on the high global searching ability and robustness of DE/rand/2/bin mutation scheme and the good performance of local searching ability and fast convergence speed of DE/best/2/bin mutation scheme, the algorithm includes several ordinary populations and a particular population. The ordinary populations take the DE/rand/2/bin mutation scheme, while the particular population takes the DE/best/2/bin mutation scheme. To balance the global searching ability and the convergence speed, some excellent individuals are migrated in proper time between ordinary populations and particular population and ordinary populations themselves. And a bi-objective mathematical model of flexible job shop scheduling is established for the sake of practical production. Finally, the algorithm is used in an example of scheduling, the simulation results show that the algorithm is feasible and efficient.