Optimal driving strategies for emergency operation of high-speed trains using on-board energy storage systems

A power outage occurs when there is an interruption to traction power system. In such emergency situations, trains are expected to achieve autonomy operation powered by on-board energy storage systems (OESS). This paper presents optimization models and methods to find optimal driving strategies for train emergency operation. Specifically, a nonlinear and non-convex program is first formulated in space-domain to minimize trip time under the limits of power and energy capacity of OESS. To improve computational efficiency, the time dynamics is removed from constraints and adjoined to the cost function. Furthermore, convex modelling steps are proposed to reformulate the problem as a convex program that can be solved efficiently. A bi-level algorithm is designed to obtain optimal driving strategies that can arrive at destinations as soon as possible. Compared to energy-optimal driving strategies, the proposed methods can effectively reduce the trip time.