为了更有效优化半导体制造系统生产资源调度,提出了基于分层着色时间Petri网模型的分时段优化调度方案.通过增强基本Petri网的描述能力和引入分层Petri网的思想,根据半导体制造系统的结构建立其改进Petri网模型-分层着色时间Petri网模型,并在模型中引入调度库所将调度算法嵌入到模型中,该模型能有效地克服基本Petri网模型规模膨胀的缺陷;同时将每一生产周期分成若干时间段,利用遗传算法来寻找各时间段内调度规则组合来优化半导体制造系统的动态性能,实现分时段优化调度,编码时,根据机器组的利用率来选择部分机器组为瓶颈机器组,重点调度瓶颈机器组,提高算法的搜索效率;仿真实验结果验证了该调度方案的优越性. In order to optimize the scheduling of production resources in semiconductor manufacturing system more effectively, a sub-optimal scheduling scheme based on layered colored time Petri nets is proposed. By improving the description ability of basic Petri nets and introducing Petri nets, Structure to establish its improved Petri net model - layered shader time Petri nets model, and introduces the scheduling database into the model to embed the scheduling algorithm into the model. This model can effectively overcome the shortcoming of scale expansion of the basic Petri nets model. At the same time, A production cycle is divided into a number of time segments, the use of genetic algorithms to find the scheduling rules in each time segment combination to optimize the dynamic performance of semiconductor manufacturing systems, to achieve optimal scheduling sub-periods, coding, according to the utilization of machine groups to select part of the machine group is Bottleneck machine group, focus on scheduling bottleneck machine group, to improve the search efficiency of the algorithm; simulation results verify the superiority of the scheduling scheme.