Passenger Flow Safety State Prediction And Risk Warning Technology Research Of Urban Rail Transit Station

Urban rail transit has the characteristics of low delay rate and small occupation of surface space.The number of passengers who choose rail transit travel increases year by year,and the pressure of passenger flow increases.As the key node of connecting network,station is the main gateway of passenger flow distribution.How to ensure the safe travel of passengers is the most important part of urban rail transit operation.To avoid the safety accidents caused by too dense passenger flow,the most effective measure is to establish an early warning system,detect the hidden dangers in advance,put forward scientific safety plans,and prevent trouble in the early stages.This paper studies the characteristics of passenger flow distribution in urban rail transit from two dimensions of time and space.In terms of time,passenger flow has periodic stability,imbalance and mutation.In space,passenger flow shows uneven characteristics from two aspects of rail transit network and station.At the same time,from the perspective of passenger flow prediction and passenger flow early warning,the demand analysis of passenger flow safety state of urban rail transit is carried out,which lays a foundation for the selection of prediction and early warning model and the verification of examples.The ARIMA forecasting model,the RBF forecasting model and the XGBoost forecasting model are selected for the passenger flow forecasting part of rail transit station.Combine the passenger flow data samples of the passages,entrances and exits,stairs and escalators of a subway station in China for prediction.On the basis of the evaluation index value of root mean square error,average absolute error,R square sum average relative error,comparing the accuracy of model prediction results,it is concluded that RBF prediction model and XGBoost prediction model have higher accuracy and are more suitable for the scene of rail transit station.Because the single prediction model has its own limitations,combined with the actual passenger flow data certificate,this paper selects the passenger flow combination prediction model from the different combinations of the main passenger flow prediction model.Based on the scope of application,the minimum square error and method are selected for the prediction model of urban rail transit station on LSSE.Combined with the passenger flow data of a subway station in China,we proved that the prediction model is more accurate than a single prediction model.According to the service level of rail transit,the design standard and the passenger flow characteristics of each area of the station,this paper divides the passenger flow warning grade into four grades,corresponding to different passenger flow states in turn.The passenger flow density,passing capacity and carrying capacity are selected as the evaluation indexes of platform,walking passage and escalator to determine the early warning threshold.Based on the real passenger flow data of a subway station in China,this paper analyzes the prediction and early warning of the safety status of rail transit stations,takes passenger flow as the research object,and improves the accuracy of passenger flow prediction by constructing a combined prediction model.Then select the appropriate evaluation index to divide the early warning grade of passenger flow.For rail transit station safe operation,to ensure comfortable travel of passengers to provide protection.