从当前半导体晶圆制造企业实际生产调度的角度出发,根据模糊控制理论和模拟退火算法,提出了解决半导体晶圆制造系统(SWFS)时变多目标生产调度复杂问题的方法.大量的仿真实验数据证明,该方法可以在求解速度和优化质量间取得平衡,并实现了全局优化效果,为实现SWFS的科学生产控制与调度提供了有效策略和方法.
This study presented a new methodology for solving time variable-multiple-objectives (TVMO) scheduling optimization problems of a complex wafer fabrication. Simulation of the virtual wafer fabrication by simulated annealing (SA) under control of TVMO scheduling is an example of the evolutionary and practical methodology. Such methodology presents a practical and effective way to solve time-consuming issue of searching feasible solution and ensures the result can meet planner's quality requested. The study of production scheduling optimization provides positive support for related researches of production planning, scheduling, dispatching and control in the semiconductor manufacturing field.