R2R controller design using T-S fuzzy model and Extended State Observer

The wafer quality metrology delay is a very common phenomenon in the semiconductor manufacturing processes. Furthermore, the metrology delay is usually stochastic and time-varying, which will affect the stability and control performance of the control system. In this paper, a new kind of Run-to-Run (R2R) controller based on both Takagi-Sugeno (T-S) fuzzy model and Extended State Observer (ESO) is developed to keep the nominal performance of system. Firstly, a R2R controller based on ESO is presented. Then T-S fuzzy model is used to modify the proposed R2R controller. Stochastic delay is analyzed using the Markov chain theory to get the membership function of T-S fuzzy model. An industrial example of a mixed-product manufacturing process is used to demonstrate the performance of the proposed R2R controller.