Scheduling Analysis Of Dual-Armed Cluster Tools For Multi-Variety Wafers Revisiting

Semiconductor manufacturing industry is the cornerstone of social informatization,digitalization and intelligent development.As the basic components of semiconductor manufacturing,the traditional mass production cannot meet the requirements of large size wafer processing.Cluster tools have a highly efficient and reconfigurable single wafer processing environment so that the harsh constraints in the wafer processing could be solved.However,the increasing size of wafers has increased the demand for multi-variety,small-batch customised products.In addition,different type-wafers are frequently switched in batches during the processing of cluster tools,which makes it more difficult to schedule and control the tools,prolonging the system production cycle and increasing the production cost.In order to improve tool production efficiency and reduce industry investment,the research on the scheduling strategy of multi-variety wafers processing will surely become the focus.In this paper,dual-arm cluster tools are selected to study the scheduling problems of multi-variety wafers with identical flow pattern,wafer residency time constraint,revisiting and mixed-processing.The main research contents are as follows:(1)The operation process of dual-arm cluster tools with revisiting and mixed-processing of identical flow pattern of multi-variety wafers is analyzed.Based on the basic definition and performance characteristics of Petri net modeling,a ROPN model for dual-arm cluster tools with revisiting and mixed-processing is established.Based on this model,a system deadlock avoidance control rule is proposed,and the system activity is obtained by analyzing the time constraints of wafer processing.(2)Based on the improved swap strategy of the robot,the local and global cycle processes of the system are analyzed,and the relationship between the robot processing time and wafers residency time in each PM is obtained.In view of different PM workload situations,an effective analytical formula is proposed to obtain the steady-state scheduling of the system,and the correctness and optimality of relevant scheduling situation theorems are verified.Thus,the minimum wafer cycle scheduling of the system is realized.(3)The 2D or 3D eM-Plant simulation model is established to describe the dynamic process of dual-arm cluster tools with multi-variety wafers revisiting,respectively.Real-time control of robot activity and type wafer processing through embedded Sim Talk 2.0 language program,and real-time statistics of relevant time points.Finally,the validity of the system scheduling strategy is verified by analyzing the residency time of each type-wafer in different PM and the system production beat.(4)In order to simplify the data interaction of eM-Plant simulation model,a Python scheduling architecture is designed to ensure the steady-state scheduling of dual-arm cluster tools with revisiting and mixed-processing of identical flow pattern of multi-variety wafers,and the integrated simulation environment principle of “Algorithm-Simulation” is expounded.The above researches show that the scheduling scheme,eM-Plant simulation model and Python data interaction theory in this paper,which can be applied to the scheduling analysis of dual-arm cluster tools and enrich intelligent simulation theory.