Research On Post-assessment For The Effect Of Urban Rail Transit Initial Passenger Flow Forcast:a Case Study In Ningbo

As one of the most important ways to alleviate urban traffic congestion,the urban rail transit is paid attention by various government departments,and more and more cities are pouring into the construction of rail transit.However,many problems have been revealed from the actual operation of many cities that have opened rail transit in China.According to the effect of many urban rail line in the initial operation,there is a prevalent problem of poor passenger fow,which is significantly lower than the forecasted passenger flow.The passenger flow forecast is an important reference for urban rail transit design and construction,the huge difference between the actual passenger flow and the predicted passenger flow will bring great resistance to the follow-up development of rail transit.Therefore,it is necessary to identify the risk factors that lead to forecast deviation,especially the initial forecast deviation,and targeted to improve the forecast accuracy,reduce the waste of resources caused by insufficient initial passenger flow.At the same time,how to take remedial measures in time to improve passenger traffic efficiency when the large deviation appears is also very important.This article first embarks from the prediction itself,on the basis of the traditional four stage method,analyze the critical factors of each stage that influence the prediction results.,and then introduces the basic connotation of post-assessment for the effect of urban rail transit initial passenger flow forcast.Besides,taking the initial operation line of Ningbo rail transit as an example to carry out the post-assessment of initial passenger flow forecast effect:compare the forecast index to find the main factors that cause the differences of Ningbo and some other cities' rail transit passensger flow,then extract the critical factors that influence initial urban rail transit passenger flow forecast precision.In the end,identify the risk factors which are difficult to control are the development of urban land use and changes in the traffic connection between the ground traffic and trail transit.Secondly,in view of the risk factors of initial passenger flow prediction,select a suitable description index,and establish the relationship model that reflect the relation between risk factors and initial passenger flow.Numerical simulations show that BP neural network model that has been optimized by genetic algorithm has strong explanatory to the relationship.On the basis of relationship model,analyze the influences of change of risk factors on the passenger flow in the initial stage.Finally,according to the research results of the risk factors of passenger flow forecast,gives some suggestions that improve the accuracy of initial passenger flow prediction and the initial passenger flow benefit to deal with the situations when initial actual passenger flow below the predicted passenger flow.