Research On Virtual Coupling Safety Protection Method Based On Train Spatio-temporal Trajectory Prediction

Virtual coupling,which has gradually become a research hotspot in the field of urban rail transit,can effectively improve the line passing capacity and the flexibility of transportation.The virtual coupling train completes the information exchange between trains through the train-to-train communication technology.The train realizes shortdistance tracking operation through collaborative control,which improves the transportation efficiency and the flexibility of the coupling may also bring challenge.The existing train safety protection methods cannot meet the needs of short-distance tracking and collaborative control of virtual coupling trains.Therefore,from the perspective of virtual coupling train operation safety protection,this paper proposes a safety protection method for trains by predicting the spatio-temporal trajectory of the preceding vehicle.method.The main work is as follows:(1)A train safety protection method based on spatiotemporal trajectory prediction is proposed.Through the complete prediction of the space-time trajectory of the preceding vehicle within the fault-tolerant time,the lower bound of the space-time trajectory prediction envelope of the preceding vehicle is used as the basis for train safety protection.And then the safety protection of the train is carried out from the two dimensions of time and space,shortening the tracking interval of virtual coupling trains as well as improving the train operating efficiency.(2)A train dynamic trajectory prediction method based on the reachable set of hybrid automata is proposed.Firstly,the hybrid automaton modeling of the train operation control logic is carried out.Then,a polygon determination rule for reachable sets is proposed to ensure the over-approximation of reachable sets.Finally,a reachable set solution algorithm based on adaptive step size is proposed,and the solution step size is adjusted in real time according to the relationship between the current state of the train and the migration target,so as to achieve rapid prediction of the over-approximate envelope of the train’s spatiotemporal trajectory.(3)A spatio-temporal train protection method based on trajectory prediction is proposed.Firstly,the protection strategy of the train is adjusted accordingly according to the train operation scenario.Secondly,the train is protected from the two dimensions of time and space to ensure the absolute safety of train protection.Finally,a calculation method of train spatio-temporal collision avoidance protection curve is proposed,which can quickly calculate the train safety protection curve by solving the safety distance required for train protection.(4)The simulation experiment of virtual coupling train operation is carried out.Firstly,the influence of calculation step size and fault tolerance time on the prediction results of train spatio-temporal trajectory is analyzed.Then,based on the real subway train operation data,the simulation results show that compared with the "hit soft wall" protection method,the total running time of virtual coupling trains can be further shortened under the safety protection method based on train spatio-temporal trajectory prediction,and with the number of trains in the formation increases,the more obvious the effect of improving operating efficiency.The safety protection method proposed in this paper is suitable for the virtual coupling train operation scenario,which is conducive to further shortening the minimum tracking interval of the virtual coupling train and realizing the efficient coordinated operation of the trains.This thesis includes 62 figures,10 tables,62 references.