Research On Short-term Passenger Flow Forecast And Congestion Control Of Urban Rail Transit Under Large-scale Events

In recent years,the rapid promotion and development of urban rail transit in various public transportation fields has become an way of travel for modern humans.At the same time,with the rapid development of my country’s social economy,more and more places have large-scale events.Before the game,a large number of people flooded into the venue,and the site near the event is bound to form congestion,posing a greater threat to the safety of passengers,and affect the rail transit operation system.Based on the accurate passenger flow prediction results of the event station,reasonable rail transit network operation and management can be carried out.In-depth study of the passenger flow changes of active stations and line networks under large-scale events through examples.This paper constructs a model predictive feature input method that combines traditional features and activity features.Using this method,an improved GA-SVM model can be constructed to accurately predict large-scale events.According to the forecast results of the short-term inbound passenger flow of the station,passenger flow congestion control measures are proposed.The main contents are as follows:(1)The study analyzes the changing laws of passenger flow in large-scale and active urban rail transit.Clarified the specific definition,characteristics and classification of large-scale urban railway activities,pointing out that they are all heterogeneous and arbitrary in time,they are also specific in space,and there are certain uncertainties in the activity crowd.The main influencing factors of the passenger flow of my country’s railway urban rail transit under large-scale incidents are summarized,the actual changes in the passenger flow of my country’s railway urban rail transit under large-scale incidents are analyzed,and the passenger flow under concentrated explosive incidents is analyzed.Research object,in-depth analysis of the temporal and spatial distribution characteristics of passenger flow in large-scale events.(2)The traditional passenger flow forecasting methods and short-term passenger flow forecasting methods are classified,the problems that should be paid attention to in rail transit passenger flow forecasting under large-scale events are analyzed in detail,and the idea of using genetic algorithms to improve the forecasting model is determined.Insufficiency,an improved GA-SVM model is proposed,and a combined feature construction method combining activity features with conventional features is proposed.(3)According to the principle of rail transit passenger flow control and the three-level passenger flow control measures,the three types of current limiting characteristics under various large-scale activities are analyzed by selecting examples.The relationship between the current limiting characteristics and control measures,and the establishment of management and control measures for the stations and surrounding areas at all levels of passenger flow.At the same time,the operation adjustment analysis of the rail transit network based on the passenger flow control of a single station is carried out,and the train routing and train stopping are proposed.Station selection plan.(4)Taking Xi’an Metro Line 2 Stadium Station inbound passenger flow prediction as a research case,an improved support vector machine model is constructed based on genetic algorithm.First,determine the input characteristics of the machine prediction model,and use parameter optimization algorithms such as particle swarm optimization(PSO)and grid search algorithms to build a combined model for prediction.Comparative analysis shows that the improved support vector machine prediction model has the best accuracy and strong applicability,and can be used to predict passenger flow of rail transit in real life.Based on the prediction results,congestion control measures for stations around large-scale event venues are proposed.