Conflict Detection Theory For Arrival And Departure Lines Of High-speed Railway Station

With the rapid development of national economy,the cross-regional allocation of resources has become more frequent,and the frequency of people’s travel has increased significantly.High-speed railway is favored by passengers because of its convenience,safety,comfort,and environmental protection,but at the same time,the punctuality of high-speed railway operations has also attracted more attention from passengers.In high-speed railway stations,the rational use of arrival and departure lines can ensure the normal transport order of the station,and is significant for coordinating various transportation resources.Faced with the complex operating environment and the interference of uncertain factors,station dispatchers relying only on manual experience to identify the conflict between the use of the arrival and departure lines will inevitably make mistakes.At this time,the relevant research on conflict detection theory on the arrival and departure line highlights its importance.Therefore,this paper will build a prediction model of the conflict duration based on the actual train operation performance data,so as to realize the automatic detection of the conflict.The specific work is as follows:(1)From the perspective of trains occupying transportation resources,analyze the conflict in the use of the arrival and departure lines,and clarify its definition;categorize the conflict and then expound its characteristics.From the perspective of time occupancy,this paper studies the generation mechanism of the conflict,and proposes a criterion for determining the conflict at high-speed railway station.Based on the generation mechanism and the actual transportation production,the influencing factors of the conflict are analyzed.(2)Based on the train operation performance data,the related indicators of train operation information and the conflict are statistically analyzed and visualized.The indicators of arrival and departure line conflict mainly include conflict frequency,conflict rate and conflict duration.The macro distribution of the conflict is shown from time dimension,space dimension and space-time dimension.(3)Exploring the relationship between the arrival-departure line conflict transfer and train delay based on the hidden Markov model.Taking Hengshan West Station as the research object,the state transition matrix,observation probability matrix and initial probability distribution of the model under different tracks and different periods are calculated.In order to study the transition law of conflicts of different intensities,the state transition matrix of arrival-departure line conflict in different time periods based on conflict intensity is obtained.(4)By analyzing the influencing factors,select actual arrival time of current train,the arrival delay of the current train,the departure delay of the predicted train in crossed station,the predicted train’s scheduled interval running time from the crossed station to the current station and the scheduled departure interval of the crossed station as model input variables,and machine learning methods(random forest model,gradient boosting regression tree model and XGBoost model)are used to build a prediction model for arrival and departure line conflict duration.After model comparison and verification,it is found that the XGBoost model has excellent prediction effect and good applicability.