Study On Probe Vehicle System Configuration Optimization Based On Traffic Information Coverage

Compared to the conventional traffic data collection methods such as roadside detector, the probe vehicle technology owns the advantages of less investment and larger range of data acquisition, while it has the disadvantages such as huge variance existed in individual data and low level of traffic information coverage caused by the uncertainty of driving behavior. However, with the increasing of probe vehicle number, the traffic information coverage of road network is improved, while the cost of initial investment and system maintenance will usually substantially increase. Therefore, it is essential to formulate reasonable configuration scheme for probe vehicle system to improve service level and reduce cost.This study aims at studying how to determine the reasonable number of probe vehicles to achieve the system configuration optimization for probe vehicle system based on the traffic information coverage. On the basis of review of related study, this study analyzes the characteristics of the probe vehicle system to determine the main evaluation indexes of the information service of the probe vehicle system. Based on analysis of probe vehicle data collected from Beijing and Shenzhen, this study explores the factors affecting traffic information,°and improves the system configure scheme by optimizing the number of probe vehicle according to elastic analysis on a specific road network. Afterword, as for the shortcomings on transferability of existing coverage model, this study proposes macroscopic traffic information coverage model using multiple linear regression method with the indicators describing road network structure as the variables. Meanwhile, through analyzing the inner mechanism of the coverage, this paper studies the influence of the link attributes such as road hierarchy and road location on the number of probe vehicles passing through the link in5minutes, and proposes a microscopic traffic information coverage model based on the maximum likelihood estimation method. The data collected in Nanjing is used to testify these two models, and the result shows that the two models both own good performance in transferability and accuracy. Finally, the system configuration optimization model is established under the constraints of the minimum traffic information coverage requirements and the contribution of vehicle number to the improvements of probe vehicle coverage. It is high expected that the proposed system configuration optimization model could provide an important reference for the system configuration scheme of a new probe vehicle system in different cities.