MFRL-BI: Design of a Model-free Reinforcement Learning Process Control Scheme by Using Bayesian Inference

Abstract

This paper proposes a model-free reinforcement learning (MFRL) process control scheme using Bayesian inference (MFRL-BI) to reduce variations in manufacturing systems. The MFRL-BI controller updates the distribution of disturbances in real-time using Bayesian inference to improve control accuracy. The effectiveness of the proposed method is demonstrated in a nonlinear chemical mechanical planarization (CMP) process.

Juan Du

Assistant Professor

My research interests include knowledge-infused data science for quality improvement, industrial data analytics and machine learning, and system informatics and control for manufacturing applications.