Route Selection for Vehicle Navigation and Control

This paper presents an application of neural-fuzzy methodology for the problem of route selection in a typical vehicle navigation and control system. The idea of the primary attributes of a route is discussed, and a neural-fuzzy system is developed to help a user to select a route out of the many possible routes from an origin to the destination. The user may not adopt the recommendation provided by the system and choose an alternate route. One novel feature of the system is that the neural-fuzzy system can adapt itself by changing the weights of the defined fuzzy rules through a training procedure. Two examples are given in this paper to illustrate how the route selection/ranking system can be made adaptive to the past choice or preference of the user.