Reinforced performance control for automatic train operation with speed measurement deviation: a recollection observer approach

High-speed automatic train operation boasts extensive development prospectives due to its operational efficiency enhancement and automation capability. However, the potential running risk arising from speed sensor faults of returning biased measured speed values has emerged as one of the primary concerns. This paper presents an innovative recollection observer method to address the challenge posed by undetected speed sensor deviations, which often stem from various operational factors. Furthermore, a new reinforced prescribed performance control (PPC) framework is proposed to stabilize the error dynamics related to target position and speed trajectories between adjacent stations of high-speed trains, thereby enhancing system performance and reliability against speed measurement deviation. Compared with the previous PPC method, the “reinforced” design incorporate the synthetic effect of direct defined errors and well-posed transformed nonlinear errors, which makes it possible that defined errors can be tuned to arbitrarily small even without decreasing boundary functions and with undetectable sensor deviation. Furthermore, rigorous closed-loop system stability analysis using Lyapunov theorem is given considering the discontinuity raised by sensor faults and piecewise gradient profiles. Experimental results by implementing the proposed control to G1 high-speed train of Beijing-Shanghai railway line, are presented to demonstrate the effectiveness.