Particle filter based traffic state estimation using cell phone network data

Using cell phones as traffic probes is a promising Intelligent Transportation System technology. Compared with traditional traffic data collecting approaches, cellular probe has the advantage of the ready-to-use infrastructure and the wide coverage. This paper presents two Bayesian framework based traffic estimation models by the measurement of cell handoff data of floating vehicles. The first and the simpler model uses traffic speed as the only state variable. The second-order model, incorporating traffic volume as the second state variable, has a two-level architecture, where macroscopic states and microscopic states are connected by the process of state reconstruction. This mechanism makes it possible to realize high-order sparse-sampling traffic estimation. Owe to the good performance on solving highly nonlinear estimation problems, particle filters are introduced to provide the approximation solution of traffic state estimation problems with system noise and measurement error. The performance evaluation and practical test of particle falters under different data sets are performed by numerical experiments