Braking-Penalized Receding Horizon Control of Heavy-Haul Trains

Incorporated with a receding horizon control (RHC) approach, a penalty method is proposed for reducing the energy wasted by braking in the operation of a heavy-haul train. The practical nonlinear model of the train is linearized to design the RHC controller. This controller is then applied to the practical nonlinear dynamics of the train, and its performances are analyzed. In particular, the main focus of this paper is on the impact of the brake penalty on the train´s performances. Meanwhile, a fence method is presented to tackle two issues. The first issue is that all the cars in a train cannot be individually controlled due to a limit on the available transmission channels for control systems in a long train. The other issue is that the RHC approach suffers from heavy computation and memory load. Simulations verified that the brake penalty presented in the design can remarkably reduce the energy consumption and in-train forces of a train without sacrificing the velocity tracking performance of the train. Simulations also verified that the fence method is essential to reducing the related computation load when the RHC approach is applied to a long heavy-haul train. Further, it is demonstrated that the fence method can effectively shorten computation time and reduce memory usage without severely jeopardizing the performance of a train.