Research On Short-term Passenger Flow Prediction Of Urban Rail Transit Based On Data Stability Processing

With the progress of urbanization in China,more and more people flow to big cities,which brings great pressure to the traffic in big cities.Urban rail transit has become the primary way to solve this problem.With the development of urban rail transit and the increase of population,the situation of its station passenger flow is becoming more and more complicated.With the development of computer intelligent network in recent years,the operation and management of urban rail transit has begun to be intelligent and automated.Including the site into short-term prediction to become one of the most important aspects of the management of urban rail transit operation,choose reasonable high accuracy intelligent forecasting model,can be effective real-time grasp the changes of site traffic,which would be helpful to the site of urban rail transit is different,different circuits which can adjust the train departure flight and passenger flow orderly organization,improve the operational efficiency of urban rail transit.This paper mainly studies the short-term complex passenger flow prediction of urban rail transit stations.The main research contents include:(1)spatial and temporal characteristics analysis of passenger flow in Beijing urban rail transit network.This paper analyzes the passenger flow characteristics of Beijing urban rail transit line from the spatial perspective.From the point of view of time,the passenger flow of the station is classified into different types.The similarity coefficient and systematic clustering are used to analyze the temporal correlation of passenger flow at different types of stations,which lays a foundation for the analysis of passenger flow data for the prediction of inbound passenger flow.(2)Pretreatment of urban rail transit short-term passenger flow stabilization.For complex station passenger flow the intrinsic time scale decomposition method is adopted for the first time,stabilizing treatment to deal with the results obtained by decomposition component spectrum analysis to grasp the original information of passenger flow change frequency,reference sample entropy and white noise to evaluate component sequence,the use of volatile strong component moving average processing,thus making the original complex traffic into a stable weight,help to forecast next.(3)Urban rail transit short-term passenger flow combination prediction model.The LSTM prediction model with good adaptability to short-term complex passenger flow was selected as the core model of the prediction.The simulated anneal algorithm and genetic algorithm were combined to establish the simulated anneal-genetic combination optimization algorithm,and the combined optimization algorithm was used to optimize the LSTM prediction model and establish the optimized combination prediction model.The incoming passenger flow of Beijing West Railway Station is selected as the verification object of this paper,and the ITD-MA-GASa-LSTM prediction model established in this paper is used for short-term passenger flow prediction.It verifies that the method proposed in this paper is helpful to the operation and management of urban rail transit.In this paper,the temporal and spatial characteristics of the passenger flow of the station are analyzed,and the correlation coefficient is used to analyze the relevant characteristics of the passenger flow changes at different stations in different periods,which is helpful to grasp the internal laws of different passenger flows.In addition,the inherent time-scale decomposition model in the field of signal analysis is used to decompose passenger flow,providing a new and effective data processing method for passenger flow prediction and combining various data processing and analysis methods,which is conducive to improving the accuracy of the prediction model and providing a certain method and theory for the operation and management of urban rail transit.