Bayesian Parameter Auto-Tuning for ADRC Based PMSM Speed Regulation

Parameter tuning for speed regulation in permanent magnet synchronous motors (PMSM) using active disturbance rejection control (ADRC) mainly relies on a time-consuming, labor-intensive, and often expert-driven trial-and-error approach. To address this challenge, this study introduces Bayesian optimization to streamline this crucial yet tedious parameter tuning process. Within this framework, we reformulate the control parameter tuning problem as a black-box optimization challenge, which is efficiently solved by using Bayesian optimization techniques. The effectiveness of the proposed Bayesian parameter tuning approach with four different acquisition functions is demonstrated through a comparative simulation study, showcasing improved control performance compared to manual tuning.