Protection-enveloped Automatic Train Operation: A Virtual Parameter Adaptation-based Finite-time Control Approach

Abstract

Automatic train operation (ATO) is a key technology to ensure the operation safety and efficiency for trains instead of drivers. By operation safety, it is meant that the state of trains, including actual position and velocity, output by ATO, are always kept under the constrained ones imposed by automatic train protection (ATP), resulting in a protection-enveloped ATO subsystem. This paper propose a finite-time control method for automatic train operation via a novel virtual parameter adaptation technique. Threefold features of proposed method are: i), it is guaranteed that the output tracking errors with respect to target running curve are always kept between protection-enveloped related prescribed performance bounds, ii), after a finite time, the position and velocity tracking error will remain in the neighborhood of zero, and iii), the novel virtual parameter adaptation technique is capable of dealing with the uncertainty in train dynamic model and complex environments. By Lyapunov stability analysis and finite time theory, all close-loop signals are kept bounded in finite-time. The effectiveness of proposed control method is demonstrated by numerical simulations.