5G-enabled Edge Intelligence for Autonomous Train Control: A Practical Perspective

Abstract

Edge intelligence (EI) is becoming one of the research hotspots among researchers, which is regarded as a potential technology to empower urban rail transit. Combined with high capacity and reliability 5G technologies, EI can perform complex computing tasks offloaded from trains within low latency. EI can provide real-time and intelligent control functions for autonomous train control (ATC) system, which will effectively improve control accuracy, shorten train tracking interval, and greatly improve train operational efficiency. In this paper, we propose a 5G and edge intelligence empowered autonomous train control system. It is functionally divided into cloud layer, edge layer and service layer. Using google Kubernetes technology as infrastructure, the edge computing system receives train states and conducts the train intelligent control model. With the aim to minimize the tracking deviations and balance the performance of punctuality, energy-efficiency and riding comfort, we adopt model predictive control (MPC) algorithm to model the autonomous train control process. The MPC based intelligent control model is carried out in the edge computing system. Extensive experiment results illustrate that our 5G and edge intelligence based infrastructure can provide reliable and real-time computing service for autonomous train control systems. The train operational efficiency can be greatly improved with our suggested intelligent train control algorithm.