关键词: 晶圆制造系统, 工业大数据, 数据驱动, 加工周期, 数据生命周期, 预测回归

Abstract: Cycle time is one of the key performance indicators in wafer fabrication system with the accurate prediction result,and the manufacturing process can be effectively scheduled.Industrial big data produced in wafer fabrication system were utilized to extract useful information,and based on that a framework of wafer cycle time prediction method was proposed.Aiming the problem of fluctuant manufacturing state and data life cycle,a dynamic update mechanism for prediction model was designed.By comparing with other common regression methods,the effectiveness of the proposed cycle time prediction method was validated.

Key words: wafer fabrication system, industrial big data, data driven, cycle time, data life cycle, prediction and regression