The Multi-train Cooperative Formation Approach For Urban Rail Transits Based On Scenarios

In recent years,urban rail transits have developed rapidly in China.,which has become the main means to alleviate congestion in large cities due to large capacity,high speed,high punctuality and energy saving.As a key technology of urban rail transits,communication-based train control(CBTC)ensures the safe and efficient operation of trains.However,with the development of China’s economy and the process of urbanization,urban rail transits are still facing great passenger traffic pressure,especially tidal passenger flows and sudden passenger flows.However,at present,the capacity of CBTC system is close to the design limit,and it is difficult to greatly enhance the transportation capacity.Therefore,the contradiction between CBTC’s balanced operation mode and unbalanced passenger flow distributions needs to be solved urgently.This thesis proposes a more flexible and efficient multi-train virtual formation technology based on cooperative control,so as to realize the match of passenger flows and train flows and mitigate the passenger traffic pressure faced by Urban Rail Traffic.In this thesis,based on the operation principles and the structure of CBTC systems,a train control system integrated with virtual coupling is constructed.Based on the artificial potential field method,the consistency control and the convergence control of multiple trains under the virtual coupling mode are realized.Furthermore,for the line characteristics and the operation characteristics of urban rail transits,the typical application scenarios of the multi-train cooperative formation are sorted out,and the coupling relationship and the basic flow between trains in different scenarios are clarified.Finally,based on the ROS(Robot Operating System)robot platform,a multi-train cooperative formation simulation and performance verification platform is built,where the control algorithm and the hardware are integrated and the system performance is visualized by real-time monitoring of robot running state.The main innovations of the thesis are as follows:(1)The multi-train formation approach based cooperative control is proposed with the application of the artificial potential field(APF)method and the Kalman filter.According to the one-dimensional characteristics of railways,an APF model based on leader-follower is developed,where the gravitational potential field function and the repulsion potential field function are built according to the relative position and relative velocity between trains.With the Kalman filter,effects of communication delays and position errors on the control performance are mitigated.Combining the train dynamics model,the rapid and stable convergence of a multi-train formation is achieved.(2)Considering the safety constraints of train control in urban trail transits,typical operation scenarios oriented to the main track,the turnout and the station are established,where communication links and coupling relationships among trains are clarified.And the detailed process of a multi-train formation is developed based on the station map.Furthermore,the feasibility evaluation and efficiency analysis of the multi-train formation are carried out for practical operation scenarios.(3)The simulation and performance verification platform based on ROS is developed.According to the characteristics of the multi-train virtual formation,this thesis realizes the combination of cooperative control algorithm and hardware,and the platform consists of four parts: the control center,the communication mechanism,the control principle and the monitoring method,where the fusion of robot running state and realtime monitoring is realized,and the performance of typical scenarios is verified.