Development of an algorithm for car navigation system based on Dempster-Shafer evidence reasoning

Abstract

GPS (Global Position System) has been widely used in car navigation system. It helps the users to determine the car position and velocity. But the unknown GPS noise characteristic makes the position information with undesirable position errors. The map matching process can improve the position accuracy and availability of the vehicular position system. In this paper, general principle of map matching is investigated according to segmentation and feature extraction, and an algorithm based on D-S (Dempster-Shafer) evidence reasoning for GPS/MM integrated navigation system was proposed. A simulation example demonstrates the effectives of the algorithm for the car location and navigation. The simulation errors of GPS and GPS/MM integrated system were presented.