Fluctuation And Forecast Of Urban Rail Transit Passenger Flow Under Rainfall And Snowfall

In recent years,traffic congestion has become increasingly serious in China,the development of rail transit has become an effective measure to solve this problem.However,with the expansion of the scale of the urban rail network,the operation and management of urban rail are facing more severe challenges.Passenger flow is an important part of urban rail transit system,the rainfall and snowfall will cause the fluctuation of urban rail passenger flow,then affect the accuracy of urban rail passenger flow prediction and the safety of urban rail operation.Based on the study of passenger flow fluctuation,this paper analyzes the passenger flow fluctuation and the fluctuating time under different rainfall and snowfall levels,and an urban rail passenger flow forecast model under the rainy and snowy weather is proposed based on LSTM network(WI-LSTM).Finally,an operational risk assessment model for urban rail transit under the rainy and snowy weather is established based on the extension matter-element method,which also provides the theoretical basis for urban rail operation management.Firstly,this paper calculates the ridership residual for quantifying the metro ridership fluctuation based on nine terms moving average ridership,take the passenger flow from Harbin Metro Line 1 Museum Station from December 2017 to January 2019 as an example,then analyzes the full-time,time-sharing passenger flow fluctuations under different rainfall and snowfall levels.Then analyzes the impact of work days,rest days,peak hours and other factors on passenger flow fluctuation,and the reason of this phenomenon.In addition,a quantitative method for the urban rail passenger flow fluctuating time under the rainy and snowy weather is proposed,and the relationship model between precipitation,passenger flow fluctuation and the fluctuating time is established by curve fitting.The results show that: Rainfall will have a negative impact on urban rail passenger flow,but snowfall will have a positive impact on urban rail passenger flow.The impact of snowfall on urban rail passenger flow is much longer than the rainfall,and the impact of snow occurs after a period of time of the snowfall.Then,based on the passenger flow fluctuation and the fluctuating time laws,an urban rail passenger flow forecast model under the rainy and snowy weather is proposed based on LSTM network(WI-LSTM).Taking Harbin Metro Line 1 as an example,compared with the SARIMA and SVM prediction model,the values of MAE,MAPE,and MRE of WI-LSTM prediction model are lower than other models,so that the accuracy and reliability was verified.Finally,through the analysis of the factors of urban rail operational risk and the coupling relationship between the risk factors,an operational safety risk assessment model for urban rail transit under the rainy and snowy weather is established.The weight of each index is determined by triangular fuzzy number,and then an operational risk assessment model for urban rail transit under the rainy and snowy weather is established based on the extension matter-element method,Finally,taking the rainy day of Harbin Metro Line 1 Museum Station as an example,the operational risk of the station is evaluated,and the reliability of the evaluation method is verified.The research results can provide some references for the operation and management of urban rail transit.