基于模糊物元模型的高维多目标FJSP研究*

为解决高维多目标柔性作业车间调度问题，提出了一种基于模糊物元模型与粒子群算法的模糊粒子群算法（Fuzzy Particle Swarm Optimization，FPSO）。该算法以模糊物元分析理论为依据，采用复合模糊物元与基准模糊物元之间的欧式贴近度作为适应度值引导粒子群算法的进化，并引入具有容量限制的外部存储器保留较优的Pareto非支配解以供决策者选择。此外，构建了优化目标为最大完工时间、设备总负荷、加工成本、最大设备负荷与加工质量的高维多目标优化模型，并以Kacem基准问题与实际生产数据为例进行仿真模拟与对比分析。结果表明，该算法具有良好的收敛性且搜索到的非支配解分布性较好，能够有效地应用于求解高维多目标柔性作业车间调度问题。

High-dimensional multi-objective FJSP researchbased on fuzzy matter-element model

In order to solve high-dimensional multi-objective flexible job-shop scheduling problem, this paper proposes a fuzzy particle swarm optimization algorithm which is based on fuzzy matter element model and particle swarm algorithm. The proposed algorithm adopts the Euclid approach degree between the fuzzy matter element and the standard fuzzy matter element as fitness value to lead the evolution of particle swarm optimization algorithm, and introduces an external storage with constrained capacity to reserve the optimal Pareto non-dominated solutions. Besides, this paper constructs high-dimensional multi-objective flexible job-shop scheduling model, where makespan, total machine load, cost, maximum machine load and crudy are all concerned. The results of simulation based on Kacem benchmark problem and actual production problem show that the proposed algorithm has good convergence and can also achieve Pareto optimal solution with an ideal uniformity, it can solve high dimension multi-objective flexible job-shop scheduling problem effectively.