Travel time prediction is very important to the development of Advanced Traveler Information System (ATIS) and other ITS applications such as route guidance system and traffic flow management. In this paper, a comprehensive approach including short-term, medium-term, and long-term travel time predictions based on the link travel time data collected by probe vehicle system is proposed. The suitable models for these three types of prediction and the conditions for the models' application are also identified respectively. Finally, the performance of the proposed approach is examined with the Automated Vehicle Identification (AVI) data, and the accuracy and the practicability are identified successfully. For the covering abstract see ITRD E134653.
Samenvatting