求解PFSP问题的多粒子群协同学习算法

Multi-particle Swarm Optimization Co-learning Algorithm for Permutation Flow Shop Scheduling Problem

针对以最小化最大完工时间为目标的置换流水车间调度问题,提出了一种多粒子群协同学习算法。该算法在协同粒子群算法的基础上,采用了精英库种群和普通种群共同进化框架,重新构造了学习交流方式。精英库种群采用改进的综合学习策略,普通种群中的每个子群采用经验指导的精英学习策略进行局部搜索。此外,还引入了精英迁移策略,促进整个种群的信息交流与协同进化。通过在不同规模问题的实例与另外两种优化算法进行比较,仿真结果表明了该算法在解决置换流水车间问题上的有效性。

Multi- particle swami optimization co -learning （MPSO -CL） algorithm was proposed to solve per- mutation flow shop scheduling problems with the purpose of minimizing makespan. Based on the cooperative par- ticle swarm algorithm, the algorithm consisted of one elite population and several nomlal populations, leconstruc- ted the learning exchange mode. With the elite population, improved comprhensive learning strategy was adopt- ed, and with each subgroup in the common population the carry on the local search. In addition, the elite transfer experience - guided elite learning strategy was taken to strategy was introduced to promote the communication and co -evolution of the whole population. By comparing different scale benchmarks with two other optimization algorithms, the simulation results show that the algorithm was effective in solving permutation flow shop schedu-ling problem.