基于规则导向的柔性作业车间多目标动态调度算法

设计了一种具有柔性资源约束的多目标集成优化方法,建立了包括最小完工时间、最小生产成本、最大设备利用率、最大交货满意度和最优人工分配在内的多目标组合优化模型;为降低模型的复杂度,抑制组合优化模型的状态爆炸效应,采用规则导向的资源调度思想,通过调整规则概率使概率大的规则被优先选中,从而"推动"搜索过程向预期目标方向移动;采用改进的非支配排序遗传算法—NSGA-Ⅱ获得不同规则概率值的Pareto解集,并结合动态规划法求解最优人员分配方案;仿真对比与算例验证,本文算法可以有效解决柔性作业车间多目标调度优化问题.

Multi-objective dynamic scheduling algorithm for flexible job-shop problem based on rule orientation

To reduce the complexity of multi-objective dynamic optimization for flexible job-shop scheduling problem (FJSP), an integrated optimization method with flexible resource constraints was proposed and a multi-objective optimization model was built, where the minimum completion time, lowest production cost, maximum equipment utilization rate and delivery satisfaction, also optimal human resource allocation were all concerned. First of all, to restrain the explosive characteristics of combined optimization model, we adopt the rule oriented resources scheduling idea and adjust the probability of rules to ensure the rules with a high probability to be selected to push the search process toward the target direction. Secondly, we adopt the improved non-dominated sorting genetic algorithm (NSGA-Ⅱ) to obtain the pareto solution sets with different probability values of the rules and use dynamic programming method for solving optimal allocation scheme. Finally, simulation comparison and numerical examples were carried out to illustrate that the proposed method could solve multi-objective FJSP effectively.