论文部分内容阅读
高度不确定的半导体制造环境迫切需要能够快速响应生产线实际状态的调度方法,为此,在充分考虑半导体生产线特点的基础上,给出了同时考虑混合加工方式、上下游负载均衡与工件准时交货需求的自适应派工算法(Self-Adaptive Dispatching Rule,SADR)。首先,介绍了SADR的算法流程;其次,将历史调度方案或仿真调度方案作为学习样本,基于免疫克隆选择算法挖掘SADR参数与实际生产状态参数及期望运作性能之间的关系模型,实现SADR参数跟随生产线运作状态的实时动态调整;最后,基于实际半导体生产线模型进行仿真验证。仿真结果表明,SADR能有效提高半导体生产线的运作性能,生产线上的总工件移动步数平均提高2%,并且在超载环境下能够获得更好的效果。
Highly uncertain semiconductor manufacturing environment urgently needs to be able to quickly respond to the actual state of the production line scheduling methods, to give full consideration to the characteristics of the semiconductor production line, given simultaneously consider the mixed processing mode, upstream and downstream load balancing and on-time delivery of the workpiece Requirements Self-Adaptive Dispatching Rule (SADR). Firstly, the algorithm flow of SADR is introduced. Secondly, the historical scheduling scheme or simulation scheduling scheme is used as a learning sample, and the relationship model between SADR parameters and actual production status parameters and expected operational performance is obtained based on the immune clone selection algorithm, so that the SADR parameters are followed Real-time dynamic adjustment of the operating status of the production line; Finally, simulation based on the actual semiconductor production line model. The simulation results show that SADR can effectively improve the performance of the semiconductor production line. The total number of workpiece moving steps in the production line can be increased by 2% on average, and the better effect can be obtained under the overload condition.