Probe vehicle sampling for real-time traffic data collection

With the advent of Advanced Transportation Management and Information System (ATMUS), much attention has been paid to probe vehicles (PVs) sampling for real-time traffic data collection. In this paper, the sample size of PVs was discussed and an analytical model was established based on the time-varying link traffic density. In this model, many factors that would affect the accuracy of probe data were included, such as location precision, road type and link length. The Chris's PVs sampling size model (E) over bar = 1-e(-alpha pL) was exactly proved to be the ideal case of ours. Moreover, for the sake of application of this model, the capacity constrains of wireless communication system was combined. Generally, the findings of this study serve as the first step to start the vehicles sampling for the real-time traffic data collection.