Abstract: To meet the global optimization requirement of Interbay material handling system in semiconductor wafer fabrication systems,a multi-objective scheduling approach for Interbay material handling system based on composite rule was proposed.By considering constraints between various transport equipment,the multi-objective scheduling model of Interbay material handling system was established.To generate composite rules,a method based on genetic programming was proposed,in which several attributes of wafer lots were encoded with four basic arithmetic symbols in binary tree structure,and the composite scheduling rules with the best fitness were obtained by using selection,crossover,mutation and d-best policy based λ-Interchange local search method.To quickly evaluate a large number of individuals,the calculation formulas of individuals fitness were derived based on From-To table.The results of discrete event simulations showed that,compared to the scheduling methods for Interbay system in literature,these composite rules generated by the proposed algorithm achieved better comprehensive optimization on average cycle time,average delivery time and average due date satisfaction rate of wafer lots,and meanwhile had better robustness.

Key words: Interbay material handling system, scheduling, multi-objective, genetic programming

摘要： 为了满足晶圆制造Interbay物料运输系统的全局优化要求,提出基于复合规则的Interbay物料运输系统多目标调度方法。充分考虑各运输设备之间的约束关系,建立了Interbay物料运输系统的多目标调度模型。提出基于遗传规划的复合规则生成方法,该方法将晶圆工件的属性参数与四种基本运算符号按照二叉树的结构进行编码,再通过选择、交叉、变异以及基于d-best策略的λ-Interchange局部搜索方法,寻找具有最佳适应度的复合调度规则。为了对大量个体进行快速评价,设计了基于物料运输表的适应度值估算方法。仿真实验结果表明,与文献中的Interbay物料运输调度方法相比,基于所提算法生成的复合调度规则能够对工件的平均加工周期、平均搬运时间、平均交货期满足率等多个目标进行综合优化,同时表现出较好的鲁棒性。

关键词: Interbay物料运输系统, 调度, 多目标, 遗传规划