关键词: 多目标柔性作业车间调度问题, 混合遗传算法, 不确定性, 鲁棒性

Abstract: A Hybrid Genetic Algorithm (HGA) that mixed discrete simulation and Analytic Hierarchy Process(AHP) into genetic algorithm for solving Multi-Objective Flexible Job-shop Scheduling Problem (MO-FJSP) under uncertainties of part's arrive time,process time and queue rule disorder was proposed.The multi-objectives of MO-FJSP were minimize make span,minimize balance index of task and maximize average busy ratio of machines,and being subjected to process constraint and machine constraint.To obtain a more robust approximate optimal solution of MO-FJSP based on consistency evaluation,HGA obtained the robust index of multi objectives by discrete simulation,and then calculated the fitness of possible solution by AHP.By comparing with relaxation genetic algorithm,the approximate optimal solution of MO-FJSP produced by HGA was more robust.

Key words: multi-objective flexible job-shop scheduling problem, hybrid genetic algorithm, uncertainty, robustness