加工时间随机可控Job Shop前摄调度研究  

Proactive Scheduling Research on Job Shop with Stochastically Controllable Processing Times

作　　者：肖世昌[1] 孙树栋[1] 杨宏安[1] 

机构地区：[1]西北工业大学机电学院,陕西西安710072

出　　处：《西北工业大学学报》2014年第6期929-936,共8页Journal of Northwestern Polytechnical University

基　　金：国家自然科学基金(51075337;51475383)资助

摘　　要：针对Job Shop环境中工序加工时间的不确定性,建立加工时间随机可控Job Shop调度问题随机模型。采用效率指标和鲁棒性指标对调度方案进行双目标评价。提出一种分层求解策略实现双目标优化,并采用嵌入最优计算量分配策略的遗传算法求解模型。仿真实验证明了所提出模型及优化算法的可行性。通过与直接采用均值-方差模型进行双目标优化得到的结果进行比较,证明了所提出的分层求解策略和算法可以获得综合性能更好的调度方案。Aiming at the uncertainty caused by the randomness of the processing times in job shop, we propose the stochastic model of job shop scheduling problem with stochastically controllable processing times （SCPT-JSP）.We adopt a scheduling method that optimizes this problem with two objectives：（1） efficiency measure;（2） robustness measure.To solve this dual-objective scheduling problem, we propose a two stage hierarchical strategy as well as a simulation based genetic algorithm （ GA） in which optimal computing budget allocation is embedded.The feasibility of the proposed model and the strategy and algorithm for solving the problem are proved with simulation experiments.According to the comparison with the optimization results obtained with mean-variance model which optimize the dual-objective directly, we find that our hierarchical strategy and corresponding algorithm can obtain the scheduling solution with superior comprehensive performance.

关 键 词：随机模型 遗传算法 鲁棒性 JOB SHOP 最优计算量分配 加工时间随机可控 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]