A comparison of supervisory control algorithms for tool/process disturbance tracking

The performance of four supervisory algorithms: segregated exponentially weighted moving average, exponentially weighted moving average, adaptive exponentially weighted moving average, and recursive least squares, are compared in Monte-Carlo simulation of an example system dataset, as well as on real industrial data from a photolithography process. Advantages and disadvantages of the algorithms are compared with specific focus on practical issues of interest to the user.