Speed Profile Tracking by an Adaptive Controller for Subway Train Based on Neural Network and PID Algorithm

Abstract

Train running performance largely depends on the controller. In order to improve the controller performance of speed profile tracking, an adaptive control method for subway trains is addressed in this paper. Considered time-varying parameters of train motion, the train model with dynamic parameters is constructed. The model free adaptive controller combined with neural network and proportional-integral-derivative (PID) algorithm is proposed to realize adaptive control. The control laws and parameter tuning laws of the proposed controller are both illustrated, and PID gain range and actuator saturation are taken into consideration. Compared with other controllers, numerical tests are provided to verify the effectiveness and stability of the proposed controller. The controller is also applied on a real line, and the results show that the proposed controller outperforms the currently used controller. More specifically, the proposed controller reduces average absolute running time error by 0.67 s, average absolute stopping position error by 0.09 m, average maximum absolute speed tracking error by 1.22 m/s, and total energy consumption by 2.73 kWh.