Short-Term Passenger Flow Forecasting Considering Congestion Propagation Characteristics

With the continuous deepening of the urban rail transit network construction process,the problems in management and operation have become increasingly prominent,and crowded passenger traffic has become the most concerned issue for the rail transit management department.Mastering the characteristics of crowded communication and real-time prediction of short-term passenger flow of the stations is of great significance for the transportation management department to timely adjust the operation strategy in time,to adopt the current limiting measures,and to guide passengers to travel,which is the key to solving the crowding phenomenon.Based on the analysis of short-term passenger flow characteristics and influencing factors,this paper selects cloud model as a means to realize short-term passenger flow forecasting,and quantifies congestion propagation as secondary probability.Then,it analyzes and summarizes the characteristics and factors of the spread of congestion propagation.Finally,the characteristics of the short-term passenger flow forecasting study,combined with the passenger flow examples of some stations in Beijing for verification analysis,the short-term passerger flow forecast considering the characteris of the congestion propagation.The main work and conclusions of this paper are asfollows:(1)Based on the analysis and comparison of different granularity time,the periodicity,dynamics,inhomogeneity,nonlinearity and uncertainty of short-term passenger flow are summarized,and the characteris of short-term passenger flow are verified combined with examples in several aspects.Taking the calculation of the peak duration as the means,the paper analysis and verifies the impact of three influencing factors,such as operation date,operation period and site surrounding land on short-term passenger flow.(2)Building a model to quantify the secondary probability.The uncertainty and dynamics of congestion propagation are analyzed with examples,and the influence of passenger flow and residual carrying capacity of the train on congestion propagation.Based on the above analysis,the SIS model is selected to study the congestion propagation.Combined with the improved SIS model,the congestion propagation law is verified by the changing secondary probability.(3)Based on short-term passenger flow characteristics and congestion propagation characteristics,the cloud model for short-term passenger flow in multi-rule is selected to predict short-term passenger flow.Next,taking Qianmen Station and Chaoyangmen Staion of Beijing Subway as an example,the prediction accuring of multiple indicators is analyzed.Result of the analysis,the prediction accuracy is high.