PI-like Adaptive Asymptotic Control for Automatic Train Operation with Protection Constraints and Applications

Abstract

This paper addresses the PI-like adaptive asymptotic control for the automatic train operation in the presence of protection constraints. Essentially different from the existing methods, the proposed method requires no priori information of operation resistances, and no linearized approximators, such as neural networks and fuzzy systems, are utilized to deal with unknown nonlinearities online. Protection constraints from the automatic train protections (ATP) are considered in an explicit way to design control algorithm for automatic train operation (ATO), which are operated independently in practice. A PI-like adaptive asymptotic control is designed not only to proposed algorithm with low computational complexity but also to consider the sufficient utilization of ATP information into the ATO systems. Experimental and comparative results in Beijing railway Yizhuang line are presented finally to demonstrate the effectiveness and advantages of the proposed method.