Delay Propagation Prediction For High-speed Railway Based On Spatio-temporal Network In Typical Emergent Scenarios

The High-Speed Railway(HSR),with the characteristics of large transportation capacity,high velocity,punctuality,environmental friendliness,et al.,With the formation of the "eight vertical and eight horizontal" backbone network,China’s HSR has entered the era of networked operation.At the same time,the continuously increasing of the HSR network scale and the operation density can result the strong nonlinear coupling between train-to-train,train-to-line,and line-to-network.The propagation of initial train delays may result the large-scale second delays in the HSR network under emergencies,which may seriously affect the smooth and safe operation of the HSR system and the demand of passenger travel.In this thesis,a spatio-temporal network model is established to study the time-varying coupling mechanism of train-line-network in the HSR network.Based on the network model,a delay propagation model is then established to accurately predict the spatio-temporal range of the delay propagation in real-time.Results are helpful for intelligent dispatching of trains under emergencies.It is of great theoretical and practical significance for improving the emergency response capability of the HSR system and ensuring the safe and reliable operation of the system.From the perspective of network operation,this thesis studies delay propagation prediction for HSR based on spatiotemporal network in typical emergent scenarios.The main contents are as follows:Firstly,in view of the strong time-varying characteristics of the HSR network topology caused by the operation of high-speed trains,a spatio-temporal network model of HSR is constructed and a method for identifying important spatio-temporal nodes is proposed.Aiming at the dynamic change of the connection state between different origin and destination points in the HSR network,a spatio-temporal network model of the HSR is established based on the complex network theory to characterize the spatio-temporal connection mechanism between different stations.The spatio-temporal connectivity is proposed to reveal the connectivity evolution law of the global network,and the spatiotemporal betweenness is designed to quantitatively evaluate the importance of spatiotemporal nodes.Simulation experiments based on practical HSR network data and train timetable data including G/D/C trains reveal the quantitative relationship and the evolution of network characteristics between train timetable and transportation performance,which enable accurate identification of spatio-temporal important trains and stations.Secondly,aiming at the problem of accurate and rapid prediction of the spatiotemporal range of delay propagation for HSR network under emergencies,a macromicroscopic delay propagation prediction method is constructed based on the max-plus algebra method.Considering interval signal constraints,running time constraints,station interlocking rule constraints,physical facility constraints,etc.,a fusion macromicroscopic delay propagation model is constructed based on scheduling technical specifications taking into account the prediction accuracy and computation efficiency of the model at the same time A double-layer network delay solving algorithm based on the breadth-first search is designed based on spatio-temporal network and physical constraint network.Taking the local network data of real HSR for example,it is shown that the proposed method can quickly and accurately predict the spatio-temporal range of delay propagation with a second-level solution speed under different typical emergencies,which verifies the accuracy and timeliness of the proposed method.Finally,in view of the spatio-temporal heterogeneity of timetable robustness under different emergencies,a robustness evaluation method for the HSR network based on the delay propagation mechanism is proposed.Considering the heterogeneity of spatiotemporal structure and the evolution mechanism of delay propagation in the spatiotemporal network of HSR,the evolution law between initial delay and associated secondary delay is studied,and a delay sensitivity index is proposed to characterize the delay propagation characteristics of spatio-temporal nodes.Then the correlation between the centrality and delay propagation characteristics of spatio-temporal nodes is analyzed.The heterogeneous emergency scenarios are designed to evaluate the indices evolution based on the proposed network robustness indices of cascading delays.Through the simulation experiment based on the real HSR network data and train operation data in a railway bureau,the results can help the identification of delay sensitive spatio-temporal nodes.At the same time,the robustness evaluation results of the HSR network under different scenarios can provide technical support for the timetable optimization and rescheduling optimization under emergencies.This thesis has 53 figures,10 tables and 78 references.