Robust Model Predictive Control for Inductively Coupled Plasma Generation with Parametric Uncertainty

This paper designs a robust multivariable model predictive controller to reject the model mismatch of the plasma generation process and realize accurate control of the electron density with a specific gas velocity. A dynamic control model for the plasma generation process is proposed based on the inductively coupled plasma generation numerical calculation, followed by model discretization and a general model predictive controller. Then, a novel automatic robust joint tuning algorithm is developed for the uncertain plasma generation process to quickly obtain the optimal tuning parameters. The proposed tuning algorithm can minimize the system's maximum output of total variations under model uncertainty. Simulation results indicate that the automatic joint tuning algorithm yields better performance than the results of the approach without robust tuning.