The Research of Dynamic Path Planning for Centralized Vehicle Navigation

A new dynamic path planning algorithm for centralized vehicle navigation has been proposed. Base on A* search, the total cost of path is divided into three components: static cost, dynamic cost and predictive cost. Static cost depends on topology of road network, dynamic cost represents traffic information in real-time, and predictive cost comes from previous results of path planning. In order to check the validity of new algorithm, a simulative program is developed with VC++ 6.0 language and MapX tool. Experimental results show that the new dynamic path planning algorithm not only can regulate the flow of traffic in the whole road network effectively with a little extra travel cost for users, but also can avoid the phenomenon mentioned in the Braess Paradox.