Improvement on integrity and reliability of vehicle positioning by a new map matching method

Many land vehicle navigation systems adopt integration systems which consist of GPS, DR and digital map. With such integration, GPS corrects DR drifting error, and DR smoothes out GPS error and continuously gives vehicle position information when GPS is unavailable. Digital map is used for displaying and locating vehicle on the street, known as map matching. With map matching, the positioning accuracy and coverage can be increased. However, most of the map matching methods are lack of integrity, which is the ability to detect the wrong map matching position, named mismatch in this paper. Systems using these map matching methods are not able to detect and recover from wrong vehicle locating without reliable GPS signal. As a consequence, the reliability of the whole positioning system is not guaranteed. The integrity and reliability are important for a positioning system. To ensure the system integrity and reliability, a new map matching method is proposed. This method applies curve pattern matching with weighted integration of multi-criteria map matching (WIMC). The WIMC is the first step to determine the vehicle location along the road, called locating position in this paper. Then pattern matching is applied to detect mismatch and determine where the mismatch starts so that the matching process can recover from mismatch. The method is tested by using collected GPS/DR data from field tests in Hong Kong. The result shows that by using this method, the reliability of map matching processing can be maintained and the error recovery from mismatch improves the performance of vehicle positioning.