改进的基于分解的多目标进化算法求解双目标模糊柔性作业车间调度问题

An improved multi-objective evolutionary algorithm based on decomposition for bi-objective fuzzy flexible job-shop scheduling problem

中文关键词  双目标模糊柔性作业车间调度  非支配解优先策略  变邻域搜索  计数器策略  MOEA/D

英文关键词  bi-objective fuzzy flexible job shop scheduling  non-dominated solution first rule  variable neighborhood search  counter strategy  MOEA/D

中文摘要

      针对同时考虑最大模糊完工时间和总模糊机器负载的双目标模糊柔性作业车间调度问题(BFFJSP), 本文 提出了一种改进的基于分解的多目标进化算法(IMOEA/D), 同时最优化最大模糊完工时间和总模糊机器负载, 其主 要特点是: 1) 采用3种初始化种群的策略; 2) 提出了非支配解优先策略; 3) 设计了结合5种局部搜索策略的变邻域搜 索; 4) 提出了计数器策略预防陷入局部解. 运用大量实例进行了算法策略分析和对比实验, 仿真结果表明, IMOEA/ D在求解BFFJSP上具有更优性能.

英文摘要

      In this paper, the bi-objective fuzzy flexible job shop scheduling problem (BFFJSP) is considered. FFJSP uses triangle fuzzy number to represent the processing time, which is more practical. An improved multi-objective evolutionary algorithm based on decomposition (IMOEA/D) is proposed to minimize fuzzy maximum completion time and total fuzzy machine workload. In IMOEA/D, a novel initialization strategy, i.e., non-dominated solution first rule (NDFS), is proposed to obtain high quality and diversity initialize population, which combine three initial rules. Variable neighborhood search (VNS) fixing five neighborhood structure is used to improve the convergence of population. A counter strategy is designed to control the beginning time of variable neighborhood search and prevent IMOEA/D from falling into local optimal solution. Comprehensive experiments are conducted to demonstrate the effectiveness of IMOEA/D and compare it with 5 state-ofthe- art algorithms. Experimental results show that IMOEA/D has strong advantages for solving BFFJSP.