Control performance monitoring for EWMA-based run-to-run control in semiconductor manufacturing processes

The monitoring technique is a key factor to assess the Run-to-Run control performance in semiconductor manufacturing industry. In this paper, a new performance index that is a minimum variance benchmark for the model residuals obtained from closed-loop data is proposed to evaluate the EWMA controller and the process model. Based on the structure of EWMA controller and feedback invariant principle, the disturbance innovations at current run are estimated from closed-loop data by an orthogonal projection of the space spanned by past outputs and setpoints. A performance index is developed by using the ratio of a quadratic form of the estimated disturbance innovations and that of model residuals. The effectiveness of the proposed methods is shown by simulation results.