An Improved Evidential Fusion Approach for Real-time Urban Link Speed Estimation

On the basis of previous researches, an improved evidential fusion approach is presented to integrate heterogeneous multi-sensor data in Urban Advanced Traveler Information Systems. The method inherits the advantages of the previous model in terms of the real-time processing feature; meanwhile, its performance is improved by adding a mechanism to evaluate the dynamic reliability of sensors. In this paper, the whole evidence reliability algorithms, which contain both the static one and the dynamic one, are provided primarily. After that, the frame and computational procedures of the new approach are given. Finally, a simulation test and a real-world data experiment are discussed to explain the advantage of the proposed method in comparison with the previous model. The real-world data of loop detectors and GPS probe vehicles were collected from an urban arterial road section in Shanghai.