Modeling And Analysis Of Passenger Flow Distribution Trend In Public Transportation Hub

With the growth of passenger flow volume,the frequent large passenger flow phenomenon has brought enormous pressure to the public transportation network.Especially in the transportation hub during peak hours,due to the complexity of hub function and structure,passenger rapidly gathered within a short time,which bring a serious test to the service ability and a big influence to the passenger flow line.In order to grasp the passenger flow distribution laws,it is necessary to do the research on the passenger distribution trend of the hub,which has significance to relief the hub operation pressure during the peak hours.This paper takes the passenger density and the passenger distribution trend in the public transportation hub as the main study object.Based on the epidemic model of biology and the vector auto-regression model of econometrics,this paper establishes the passenger distribution trend model of the public transportation hub,and verifies and analyzes the reasonability and availability of the model.Based on the model,this paper presents the concept of real-time passenger flow control.Firstly,through the survey of typical public transportation hub,it can obtain the passenger density distribution situation of some areas of the hub during peak hours,and analyze the passenger flow distribution laws of some key locations.Secondly,based on the epidemic model,through the discretization of the traditional model,this paper establishes the vector expression form of the passenger flow distribution trend model.By using the vector auto-regression model of econometrics,introducing a stochastic term that follows the Gaussian distribution,and using the Eviews statistical software to fit each parameter in the model,this paper gives the mathematical expression of low level lagging passenger distribution trend model.By using MATLAB to simulate the trend prediction,the reasonability and availability of the model can be verified.Thirdly,by using the Gaussian process regression algorithm,this paper realizes the high level parameter fit of the passenger distribution trend model,and predicts the passenger density trend.Through the compassion of the predict value and the actual observed value,the availability of this algorithm can be verified.Lastly,based on the passenger distribution trend model,through the analysis and summary,this paper presents the concept of real-time passenger flow control of the hub.The expected results of this paper can help the public transportation hub to make the passenger transport organization programs and passenger flow control measures,which can reduce the hub operation pressure that comes from the passenger congestion,and can improve the operational organization efficiency and overall security of the hub.