在半导体制造过程中,调整设备相关参数以提高晶圆质量的同时又会造成换模时问的增加,从而导致产品生产周期的延长。为了有效地解决此类问题,对带换模约束的非等效并行机多作业族调度方法进行了研究。在描述问题域的基础上,试图在设备参数调整次数与换模次数之间寻求一个均衡点,以实现系统最小Makespan为调度目标,建立数学规模模型,构建了求解此类并行机调度问题的改进型多态蚁群算法,并对提出的调度算法进行了仿真实验分析。结果表明所构建的算法是有效、可行的。
In semiconductor manufacturing processes, the relevant parameters of equipment was adjusted to improve the high wafer quality and increase the setup time, which leaded to delay on production cycle time. To solve these problems, a multi-job families scheduling method of non-identical parallel machines was explored with setup time constraints. On the basis of describing a scheduling problem domain, a balance point between the number of setups and the number of parameter adjustments was explored to achieve the objective of minimum system Makespan The corresponding mathematical programming model was built. A modified polymorphic ant colony algorithm was struc tured to solve the scheduling problerru Simulation experiments results showed that the developed algorithm was valid and feasible.