Forecasting Passenger Flow Of Inbound And Outbound For Urban Rail Transit Extension Line

Urban rail transit has a large investment and a long construction period.At the same time,urban traffic congestion is becoming more and more serious.Residents need to travel urgently.It is urgent to use rail transit as a fast,large volume and low pollution to travel.Therefore,urban rail transit often adopts the strategy of "building by stages and opening by stages",in order to reduce the pressure of financing and accelerate the pace of construction.With the opening and operation of extension line,the passenger flow distribution of the whole line changes.Sometimes it is necessary to adjust the original dispatching plan,and the data of passenger flow is the basis for the plan.However,there is no historical data on the extension line,so passenger flow prediction is needed.In view of this situation,this paper takes the East Extension Line of Nanchang Metro Line 1 as an example to carry out the passenger flow forecast of the urban rail transit extension line.Firstly,this paper reviews the direction and methods of passenger flow forecasting in urban rail transit,and analyses the applicable conditions of different methods,so as to provide theoretical basis for choosing appropriate methods.In view of the formation mechanism of passenger flow and its characteristics of spatial and temporal distribution,this paper takes Nanchang Metro Line 1 as an example;at the same time,combined with the nature of urban land use,the land around Nanchang Metro Line 1 is classified;combined with the distribution characteristics of passenger flow and the land use situation along the line,the relationship between passenger flow and land use is analyzed.Secondly,the preparatory work of passenger flow prediction is carried out.The first step is to determine and revise the attraction range of rail transit stations,classify the stations according to their location,determine their corresponding attraction range,and then revise the range according to the characteristics of road network.The second step is to collect the location information of stations and transfer it through the Baidu Map.The longitude and latitude coordinates of the electronic map can be used to collect the information of the interest points around it.The third step is to collect the information of the interest points around the site by Baidu map,including the population information and land use.Finally,the passenger flow data of the existing stations of Metro Line 1 are collected.Then,the model of BP neural network for passenger flow forecasting is built,the number of population around the station and the information of interest points are taken as the input of the network,and the inbound and outbound passenger flow of the station are taken as the target output.The number of input and output nodes of the neural network is determined,and the optimal number of intermediate neurons is selected to establish a single hidden layer BP neural network model.The training samples and test samples of the network are composed of the 24 existing stations.Through the test and verification of the network,the prediction errors of inbound and outbound passenger flow are 7.6% and 1.6% respectively.The population number and interest information around the East Extension Station are used as network input to forecast the passenger flow.Finally,aiming at the unbalanced distribution of inbound and outbound passenger along the line,this paper adopts the traffic strategy of Long-Short Routing,establishes a mathematical model to determine the location of the optimal turnaround station for small routes based on the predicted passenger,and rationally allocates the transportation resources of the whole line according to the spatial distribution characteristics of passenger.