Efficient controller parameter tuning for a system with disturbance

It is challenging to tune the controller parameters to achieve optimal system performance. There are many heuristic methods proposed to solve the parameter tuning problem. However these methods are not satisfactory because they have difficulty to balance different system performance indices such as response time, steady state error and overshoot. Moreover, it is hard for these methods to achieve desired system performance due to system complexity, noise and uncertainties etc. This paper proposes a parameter tuning method based on Gaussian Process Regression surrogated Bayesian Optimization Algorithm (GPRBOA). The proposed method is applied to optimize the PID parameters by iteratively constructing a system model and predicting the optimal control parameters. A servo motor position control system with payload disturbance is developed to validate the proposed method. Because proportional-integral-derivative (PID) controller is widely used in industry process control, it is used to control the servo motor. Experiments were performed and the results demonstrate the effectiveness of the proposed method.