With significantly improved positioning accuracy and coverage, trajectory data collected from GPS equipped probe vehicles has great potential for traffic speed estimations. This paper presents an innovative algorithm for estimating the traffic speed on the urban road network based on GPS data. The authors firstly adopt the GIS digital map as the foundation of mapping the GPS trajectory data to the traffic network. Secondly, a novel synthetic algorithm for traffic speed reconstruction is proposed, which assign dynamic weights to GPS records according to their different velocities and take into account the spatial and temporal relevancy. The complex relevancy of speed across different segments and time periods was identified using an Artificial Neural Network (ANN) model. At last, the proposed methodology was evaluated by comparing estimation speed with ground truth obtained from fixed detectors in Zhangzhou, China. The experiment results clarify that the proposed algorithm is effective and efficient in reconstructing the traffic state of urban road networks using GPS equipped probe vehicle data.

CICTP