Online optimized parameters for RLS-LT controller - application to processes with mixed products and feature aging

Abstract

This paper presents an iterative approach, which provides optimal solutions of the best achievement performance in the case of closed-loop aged-based processes applied on semiconductor manufacturing industry. The aim is to provide an online parameter estimation of RLS-LT controller in the case of mixed product Run-to-Run (RtR) control. The online estimation is done without any a priori knowledge of process parameters and noise structure. The best achievement performance improvement is illustrated using a simulation problem from literature and the example is applied on the removal rate control in a Chemical Mechanical Polishing (CMP) process.