Prediction Of Traffic Flow At Railway Station Entrances Based On SVM And LSTM

In recent years,with the continuous development of society and economy,the exchanges between cities have become more frequent,resulting in an increasing number of people.The railway station urgently needs a plan for timely predicting the traffic flow of the station in the future,and optimizes the structure of station service personnel in advance.Railway station passenger flow forecasting is one of the most challenging problems in the field of intelligent transportation systems.It is a key hub for preventing and alleviating railway station traffic congestion.However,there is currently no effective solution for railway station customer flow forecasting in China.The main method is still the statistical method,that is,the differential autoregressive integrated moving average model(ARIMA).This method is difficult to determine the model parameters,and the modeling is difficult,resulting in low prediction accuracy and large delay after modeling.Therefore,this paper proposes a prediction method of traffic flow at train station entrances based on Support Vector Machine(SVM)and Long Short-Term Memory(LSTM).In the method,traffic flow data is timely obtained through the front-end video surveillance device at the train station entrance and subsequently extracted and uploaded to the database server.The traffic flow data is cleaned,filtered and organized through the terminal server and is trained,learned and modeled respectively through SVM and LSTM algorithm according to the historical traffic flow data in the train station database.The model is used to predict the traffic flow at the railway station in the future,then the server retrieves decision base to make intelligent decisions based on the passenger traffic forecast information,and informs beforehand relevant staff of the decision scheduling scheme to prepare.This method can effectively reduce labor costs,save equipment resources,improve work efficiency,and improve passenger experience.At the same time,it also improves the level of railway station automation,information,and intelligence.Through experiments,it is proved that the method of predicting the traffic at the entrance of a railway station based on SVM and LSTM can effectively learn the complex time characteristic relationship in large sample data with time series,predict the traffic flow at the railway station,provide strong decision support for the management of the railway station,make the correct scheduling plan,optimize the structure of service personnel,and speed up the intelligent construction of railway stations.