CLSP问题的周期调度方法及其PSO求解

Cyclic approach for capacitated lot scheduling problem and its PSObased algorithm

采用分解思想考虑多阶段CLSP问题,从多阶段生产系统抽象出单阶段生产环节,提出以周期方式对该生产环节进行生产批量调度。在对CLSP周期调度问题进行描述和界定的基础上,建立了相应的数学模型,讨论了周期调度方法中的周期上界以及周期长度与物料批量大小之间的关系等性质,采用基于三层编码的粒子群优化算法进行问题求解。源于冷轧生产实际的计算实例表明周期方法能够大大降低问题的规模且所得设备调整费用比人工方法减少约16%。

By considering multi-stage capacitated lot scheduling problem with a thought of decomposition and abstracting several single stages from a multi-stage manufacture system,this paper proposes a cyclic approach for the capacitated lot-sched-uling problem on each abstracted stage.The corresponding mathematical model is built based on the problem description and statement,and some properties of the cyclic approach,such as upper bound for each cycle and the relationship between cycle length and lot sizes,are discovered.A PSO-based algorithm is employed with three-dimensional solution presentation.By computational tests with a real case from a cool rolling plant,it can be proved that the cyclic approach leads to a dramatic reduction of problem space and makes the setup costs degrade 16%.