Resilience Analysis And Optimization Of Train-wayside Communication Systems In Communication-based Train Control Systems

In Urban Rail Transit systems, Communication-Based Train Control(CBTC) uses continuous bi-directional train wayside communication system to achieve the information exchange between trains and wayside equipment and to ensuire the safe and effective opreration of urban rail transit system.CBTC is a safety-critical system with high requirments for availability and reliability of train wayside communication system providing data communication service for safety-related information. Since the disturbances in wireless communication system such as channel fading, wireless interference and wireless handover would affect the performance of wireless communications, the Quality of Service (QoS) of train wayside communication system will decrease due to the disturbances in wireless communications during the motion of trains, which will lead to the increase in transmission delay or packet drop rate and will affect the punctuality and accuracy of CBTC data transmission. The decrease of train wayside communication system QoS leads to the delay of control information and state information in signaling systems,which will result in unplanned service braking or emergency braking and the decrease in urban rail transit efficiency. Therefore, improving the reliability of train wayside communication system,reducing the decrease in train wayside communication QoS due to disturbances, and recovering QoS level after disturbances have become the main problem in CBTC train wayside communications.In this thesis, we study the train wayside communication system in CBTC systems.Resilience of train wayside communication system is introduced as system property to describe the preservation and recovery of train wayside communication system QoS under disturbance in wireless environment. Based on the analysis of train wayside communication system resilience, the control strategies conbining communication and train control will increase train wayside communication system resilience and further improve the smoothness of train operation as well as efficiency of urban rail transit system.This thesis analyzed the train wayside communication requirments of CBTC systems and the necessity to study train wayside communication system resilience.Based on traditional wireless communication QoS measure such as transmission delay and packet drop rate, the theoretical framework of train wayside communication system resilience is established to analyze the conservation and recovery of transmission performance after disturbances such as channel fading, interference and handover. Then the unplanned traction phase due to train control information delay is analyzed, as well as the relationship between train wayside communication system resilience and train operation smoothness. A CBTC train control model considering transmission delay and packet drop in wireless information transmission is proposed to perform quantified research of relationship between train wayside communication system resilience and CBTC system performance. In order to reduce the impact of train wayside communication system on CBTC system performance, two control methods based on Cognitice Control Method were proposed in this thesis. The cognitive control model of CBTC system using Wireless Local Area Network (WLAN) and Long Term Evolution for Metro (LTE-M) are established. Reinforcement learing was used to generate optimal control strategy considering both wireless communication parameter adaption and train control strategy to improve train wayside communication system resilience and CBTC system performance.The innovations of this thesis are as follows:1. The basic feature of resilience was introduced from interference, performance and measure. Based on the distinct features of CBTC systems, the impact of train wayside communication QoS on train tracking performance in CBTC systems was analyzed. The necessity of introducing train wayside communication system resilience was then described. The theoretical framework and quantified performance measure of train wayside communication system resilience were established.2. The state space model of CBTC train control systems was established to calculate the impact of train wayside communication system resilience on CBTC system performance. Transmission delay and packet drop rate were considered in state space model as wireless communication disturbances, and the stability of state space model with disturbances was also calculated. Simulation results showed that the better train wayside communication system resilience reflect on better train operation smoothness and higher train operation efficiency.3. A cognitive control method based on MAC-layer Packet Retry Limit Adaption(PRLA) is proposed to increase the resilience of WLAN-based train wayside communication system as well as CBTC system performance. Cognitive control methods are introduced in this thesis to improve both information transmission performance and train control performance. In cognitive control model, Information Gap,which reflects the difference between current information and sufficient information for optimal control, was defined as train wayside communication system resilience to describe train control information delay. Q-learning method was applied to calculate the optimal cognitive control strategy which chooses the optimal MAC-layer packet retry limit to improve both train wayside communication system resilience and CBTC system energy efficiency. Simulation results showed that the proposed PRLA-based cognitive control scheme can significantly improve the resilience of WLAN-based train wayside communication system as well as the smoothnessof CBTC train operation.4. A cognitive control method based on resource managementfor LTE-M based CBTC system was proposed to both improve train wayside communication system resilience and CBTC system performance using Device to Device (D2D)communication. This thesis established the train-centric train control system based on LTE-M. The data transmission process through both LTE-M cellular link and D2D link were analyzed. The cognitive control model of LTE-M based CBTC systems was given and the imformation gap in cognitive control model was defined as train wayside communication system resilience based on both LTE-M cellular link and D2D link.Q-learning method was used to calculate the optimal resource management strategy in order to opitimize the joint objective function of information gap and CBTC system performance. Simulation results showed that the resource management based cognitive control scheme proposed in this thesis can improve both train wayside communication resilience and train operation smoothness for CBTC systems based on LTE-M.