用MC-QPSO算法求解并行流水车间调度问题

Parallel flow-shop scheduling problem based on cooperative evolutionary quantum particle swarm optimization algorithm with multi-populations

针对并行流水车间调度问题的特点,提出了一种基于多种群协同进化的改进量子粒子群算法(MC-QPSO)进行求解。首先将整个量子粒子种群分解为多个子种群,然后各个子种群独立地演化,并通过周期性共享搜索信息,以获得对自身信息的更新。最后,通过具体仿真实例进行了求解验证,结果表明,在求解并行流水车间调度问题时,基于多种群协同的量子粒子群算法,在收敛速度、寻优性能等方面,都要优于遗传算法。

According to the characteristics of parallel flow-shop scheduling problem,a new quantum particle swarm optimizer,called the cooperative evolutionary QPSO with multi-populations（MC-PSO）,is presented based on the analysis of the standard QPSO.The whole quantum particle swarm group is divided into several sub-groups.Every subgroup evolves independently and updates sharing information periodically.This paper uses a practical analysis to confirm the performance of the method.The results show that MC-QPSO is effective in solving the problem.The results of simulation indicate that MC-QPSO performs better than the genetic algorithm.