Study On The Short-term Passenger Flow Forecast Model Of Urban Rail Transit Festival

The rapid economic development has promoted the increasingly perfect urban rail transit network.At the same time,the continuous improvement of people’s living standards has also led to an increasing demand for urban rail transit travel during the holidays.Between the carrying capacity of urban rail transit lines and the passenger demand during the holidays There is a mismatch,which not only puts pressure on the operation and management of urban rail transit,but also makes it difficult to effectively ensure the efficiency of passenger travel.Therefore,it is necessary to predict the passenger flow of urban rail transit during holidays and adjust the operation plan for urban rail transit during holidays.Provide basic decision-making basis to ensure transportation safety and improve passenger travel efficiency.First,the thesis analyzes the holiday passenger flow through the processed AFC data of Xi’an Urban Rail Transit.After preprocessing the AFC data of urban rail transit,the holiday passenger flow characteristics are analyzed and summarized from the three aspects of the entire network passenger flow characteristics,line passenger flow characteristics,and station passenger flow characteristics,and the key research stations are identified by this,and the research is focused on the focus The passenger flow during station holidays will provide a basis for the short-term passenger flow forecast below.Secondly,considering that there is less historical data on holiday passenger flow,and more historical data needs to be input when making predictions,these historical input data are not all suitable for historical data on holiday passenger flow prediction,and historical data at different time granularities The similarity and stationarity are different,so for the historical input data of different periods selected,under the different time granularities of 15 min,30min,and 1h,they are tested for similarity and stationarity to determine whether the historical data of different periods can be As the input data of the prediction model and make recommendations on the best prediction time granularity.Then,the station’s holiday short-term passenger flow model was constructed.Firstly,on the basis of the aforementioned research,we chose to analyze the principles of ARIMA(Autoregressive Integrated Moving Average Model)and SVR(Support Vector Machine for Regression)prediction models,combined with the analysis of the similarity and stability of urban rail transit,and proposed ARIMA-SVR combination model predicts short-term passenger flow during holidays.After comparing and analyzing the prediction results of the three models,the parameters are continuously optimized and iterated to obtain an excellent ARIMA-SVR combination prediction model;then the station’s short-term passenger flow is predicted by the ARIMA-SVR combination prediction model.Taking the Xi’an urban rail transit network as an example in 2018,the three models of ARIMA,SVR,and ARIMA-SVR are used to predict the passenger flow in and out of the first day of 17 major research stations in Xi’an in 2018,and the three models are compared and analyzed.Prediction results,constantly optimizing the iterative model parameters,to obtain excellent ARIMA-SVR combination prediction,and verify the applicability of ARIMA-SVR model,so the prediction result of ARIMA-SVR model can provide a certain reference for urban rail transit operation management.