基于萤火虫算法的学习遗忘效应 BFSP 问题研究

摘要

对于以最小化最大完工时间为目标的阻塞流水车间调度问题（BFSP），现有研究较少同时考虑学习效应及遗忘效应对生产调度的影响，为此构建了 BFSP 问题的学习遗忘调度模型，结合基于Pairwise 的局部搜索策略，应用萤火虫算法对小批量生产时的学习遗忘效应 BFSP 问题进行求解。对 Car 类问题及其学习遗忘调度模型的大量仿真测试，表明了改进萤火虫算法求解该类问题的可行性和有效性。同时，证明了学习效应能够降低最大完工时间，从而提高生产效率；而遗忘效应会使得学习效果减弱，从而导致最大完工时间的增加，学习效应和遗忘效应在生产调度中的影响是客观存在且不可忽略的。%Focusing on the issue of the blocking flow shop scheduling problem (BFSP)where the objective is to minimize the maximam makespan,it is found that present researches have less concerned the learning effect and forgetting effect in production scheduling simultaneously,so,a scheduling model considering learning and forgetting effects of BFSP was established.Combining with the Pairwise-based local search,firefly algorithm was applied to solve the blocking flow shop scheduling problem of small batch production with learning and forgetting effects.Firefly algorithm was tested by Car instances and their scheduling model considering learning and forgetting effects.The massive simulation results indicate the feasibility and effectiveness of the advanced firefly algorithm. Learning effect can reduce the makespan and thus improve the production efficiency while forgetting effect will weaken the learning effect so as to increase the makespan.The influence of learning and forgetting effects in production scheduling is of objective existence and can’t be ignored.