遗传-分布估计算法求解化工生产中一类带多工序的异构并行机调度问题

摘要

A genetic algorithm-estimation of distribution algorithm (GA-EDA) was proposed to optimize the makespan criterion for a kind of heterogeneous parallel machine scheduling problem, i.e., the heterogeneous parallel machine scheduling problem with multiple operations and sequence-dependent setup times (HPMSP_MOSST), which widely existed in chemical production. Firstly, a probability model training mechanism based on GA was presented and used to increase the information accumulation of the probability model at the initial stage of the evolution, and then the efficiency of search was improved. Secondly, an effective hybrid strategy of GA and EDA was designed to help the algorithm achieve a reasonable balance between global exploration and local exploitation abilities. Computer simulation showed the effectiveness and robustness of the proposed GA-EDA.%针对化工生产中广泛存在的一类带多工序的异构并行机调度问题，即部分产品需多工序加工，同时不同产品间带序相关设置时间的异构并行机调度问题(heterogeneous parallel machine scheduling problem with multiple operations and sequence-dependent setup times, HPMSP_MOSST)，提出了一种遗传-分布估计算法(genetic algorithm-estimation of distribution algorithm, GA-EDA)，用于优化最早完工时间(makespan)。首先，提出了一种基于GA的概率模型训练机制，用来提高概率模型在算法进化初期的信息积累量，进而提高搜索的效率；其次，设计了一种有效的 GA 与 EDA 混合策略，使得算法的全局探索和局部开发能力得到合理平衡。计算机模拟验证了GA-EDA的有效性和鲁棒性。