Modeling system dynamics of interacting cruising trains to reduce the impact of power peaks

Abstract

Rail operators around the globe are striving to improve the efficiency, automation, safety, and sustainability of railway systems. Despite significant advances in technologies such as artificial intelligence and automated train operations (ATO), achieving these goals is challenging for complex rail networks when accounting for unpredictable factors that alter real-time operations. In this paper, we model railway traffic in a corridor as a string of interacting cruising trains, each subject to random speed variations that are described by a stochastic process. We simulate this dynamic system under assumptions that model human drivers and ATO systems, and compute performance measures focusing on energy consumption and the power peaks arising when multiple trains accelerate simultaneously. Different strategies to smooth these peaks are investigated, including the use of regenerative braking energy, potentially coupled with an electric energy storage, and a rule that uses fixed waiting times before re-acceleration. Our findings shed light on when and why these strategies can be effective at reducing energy consumption and/or shaving the peaks. They also show that employing a well-calibrated ATO controller in which vehicles exchange information about their location improves energy performance compared to a model of a human driven. Finally, a trade-off between energy performance and traffic regularity is exposed, i.e., strategies to reduce power peaks may slow rail traffic down, reducing capacity utilization.