A novel dynamic path optimization method for urban traffic networks

Aiming at the problem that traffic flow varying with the departure time has not been taken into account in static path optimization method, a path optimization method based on traffic flow prediction is proposed. First of all, a novel road-based model, based on the characteristics of traffic network and the human's understanding to the road traffic network, is constructed. Then, a prediction method based on single time series is adopted to obtain the future information of traffic flow of road sections, and transportation cost of travelers for each section is estimated by the forecast result. Next, the change of road traffic situation during the traveling from the origin to the destination is integrated into path search process of improved A* algorithm by the estimated transportation cost, thus a high efficient dynamic improved A* algorithm for path optimization is obtained. Finally, the path-search examples show that the predicted traffic flow and measured traffic flow can march well, thus the path searching result can guide travelers to improve their driving efficiency.