Study On Optimal Control Of Communication-Based Train Control(CBTC) Systems Based On Game Theory

Communication based train control (CBTC) system adopts the train-ground communication to achieve the high capacity, continues and bidirectional data exchange between trains and the control center, which can achieve precise closed-loop control. Wireless local area network (WLAN) is applied in the train-ground communication of CBTC. However, there exist transmission errors in WLAN such as delay and packet drop, which may cause the delay or drop of control commands and lead to unnecessary traction and braking. Therefore, the performance of CBTC is affected.The thesis aims to solve the optimal control problem of multiple trains in the transmission error of train-ground communication. Treating the CBTC system as a network control system (NCS), the thesis establishes the model of multi-trains decision processes through game theory and proposes a game-theoretic optimal control method to decrease the tracking error, velocity error and energy consumption, which can improve the performance of CBTC.The specific research contents of the thesis are as follows.1) The thesis researches the link layer protocol of WLAN and studies the theory of NCS and game. The delay and packet drop in train-ground communication of CBTC are analysed. The thesis also proves the feasibility of applying two methods in CBTC.2) The thesis studies the method of modeling the train-ground communication and train control together through NCS. Treating the CBTC system as a network control system, the thesis quantifies the relationship between train-ground communication and train control.3) The thesis establishes the model of multi-trains decision processes of CBTC through game theory. The thesis analyses the game type of multi-trains decision processes in transmission error, defines the player, strategy space and the payoff function in multi-trains game and analyses the solution of the game. The thesis proposes a game-theoretic optimal control method, decreasing the tracking error, velocity error and energy consumption.4) The thesis verifies the optimization of the proposed game-theoretic optimal control method. The thesis determines simulation parameters and designs the algorithm of the proposed method. According to the simulation results, the performances of current control method and proposed control method are compared, verifying the optimization of the proposed method.The simulation results show that, compared with the current control method, the proposed method can decrease the tracking error by27.47%, velocity error by47.65%and save energy consumption by6.34%when the packet drop rate is0.01. The proposed method can decrease the tracking error by5.64%, velocity error by20.79%and save energy consumption by12.35%when the packet drop rate is0.1.