SHORT-TERM TRAVEL TIME PREDICATION FOR URBAN ROUTER

With the rapid development of various technologies and equipment of location based services, predicting the motion laws of moving objects based on traffic networks has become a hot research field. Advanced travel information systems and route guidance systems are an important part of the traffic operation and management. As a key parameter, travel time is the most important index to identify traffic state and the most direct evidence for travellers to make travel decisions. For the issue of the low accuracy of the travel time prediction, the paper introduces a method for travel time predication with multi-source data fusion. The method combines the technique of travel time prediction with pattern matching to create traffic pattern rules with multi-source data fusion. Simulation experiments for vehicle navigation systems with multi-source data fusion prove that the travel time predication method based on multi-source data fusion can effectively lift the accuracy of travel time prediction.