Nowadays, more and more probe vehicles are used for traffic state estimation. However, some probe vehicles don't provide positions as accurate as global positioning systems (GPS), but they can provide useful location-specific information for traffic state estimation. The resulting topological position (TP) data with accuracy relative to a segment or area, fail to provide travel distance that is necessary for traffic speed estimation. Considering the wide availability of TP data in the existing communications network, this paper proposes a method in the form of an algorithm which makesbest use of low accuracy position data for traffic state estimation. Thismethod, which is based on a stochastic framework, is quite tolerant of inaccuracy and can serve as a platform for data fusion with other data sources. Validation through simulation, described here, shows that this method is robust, can correct large bias and diminish random errors.
Abstract