Research On The Influence Factors Of Private Cars Route Choice Based On RFID Data

With the development of social economy and the acceleration of urbanization and motorization,the traffic congestion has become an increasingly serious problem.Traffic demand management as an effective measure to solve the traffic congestion has important application value in developing countries.Correctly analysis the vehicle's route choice behavior can help analysis the regularity of spatial and temporal distribution of traffic flow.In addition,it lays the foundation for network traffic assignment and implementation of demand management..Based on the analysis of drivers' route choice behavior and its influencing factors,this article used actual travel path data of private cars and alternate path data in Nanjing to construct the route choice model,and finally completed the quantitative analysis of factors' influence on route choice and the relationship between factors.The main research contents of this article are as follows:Based on the RFID path data in Nanjing,this article used the database query to select and clean the original data,converted the concept of location information,and cut the path.By constructing the road network model and applicating the Floyd shortest path algorithm to fill the missing path data,this article can obtain private cars' trajectory data.Combining the background knowledge of road traffic,it can eventually get the drivers' characteristics,travel characteristics and external conditions.Based on the road network model of Nanjing,this article applied deletion algorithm to calculate the K shortest paths.Then some rules were set to eliminate path which beyond the critical threshold,or is not in conformity with the actual road conditions,or other unreasonable path,such as loop.Finally,the optimal path under different target was got to constitute the limited and reasonable alternative paths.Based on the theory of utility maximization,the MNL model,PSL model and PSCL model based private cars route choice model were developed.This article applied the relative impedance to eliminate the effect of the dimensions of different impedance variables to the utility.And correlation analysis was apllied to eliminate the factors which are little correlation with path selection.The model parameters were calibrated by TransCAD software.Comparative analysis,elastic analysis and the marginal rate of substitution analysis were used to analyse the result of the model.The results show that it will improve the precision of the model and optimize the structure of the model using relative impedance.The PSCL model performs best in model fit and predication rate under relative impedance,and it's the most effective model in the three models.Private car drivers incline to choose the route with less travel time,shorter travel length,lower frequency of the left turns and right turns,higher total ratio of expressway and lower total ratio of spur track.Among these factors,the changes of travel time and travel length have the significant influence on drivers' route choice.Meanwhile,as to the private car drivers in Nanjing,the utility of 1 kilometer is equivalent to 4.85 minutes,the utility of 1 left turn is equivalent to 2 right turns,and if the total ratio of spur track has increased 10 percent,the total ratio of expressway must have increased 18 percent to guarantee the same path choice probability.These research results can provide the basis for formulate corresponding traffic management measures,in addition,the proposed RFID data processing method and the modeling approach of route choice behavior are of positive significance to improve the use value of RFID data.