An Moblie Edge Compting Based Train Control System With LQG Control

Nowadays, the construction and development of urban rail transit network is facing many challenges. With the increasing demand for transportation capacity, the shortage of on-board computing capacity has become one of the important problems. In recent years, edge computing (EC) has become a new solution. Through edge computing, complex on-board control computing tasks can be unloaded to the trackside computing server, and the computational power of on-board equipment can be liberated. The fifth generation (5G) mobile communication has the advantages of high capacity and high reliability, and combining 5G with edge intelligence (EI) can show lower delay and higher stability in train control systems. In this paper, we propose an autonomous train control system with edge intelligence and 5G.The system is based on Google Kubernetes high reliability edge computing platform to implement automatic train control algorithm, which is divided into infrastructure layer, edge computing layer and cloud service layer in function. The linear quadratic Gaussian control (LQG) algorithm is used to model the train autonomous control process, and the cloud safety computer is used to ensure the high reliability of the whole system without being affected by local faults. Extensive experimental results show that the train control system proposed by us based on 5G and EI has high operational reliability and data security, as well as low delay performance in the communication packet interval. Using this train control system can greatly improve the efficiency and safety of train operation, so as to improve the quality of train operation.