Forecast Map Matching Model for Vehicle-borne Navigation Based on Roadway Characteristic

Industry of vehicle-borne navigation develops rapidly in recent years. The navigation can be realized only when the positioning information gets in touch with digital navigation map. Map matching technique is the bridge of positioning information and digital navigation map. Map matching technique can determine which road the vehicle runs on by the positioning information, and the vehicle icon can also be displayed on the digital map. To study an efficient, timeliness and high precision algorithm is very important for an excellent navigation product. General map matching method is based on simple point projection. It often appears matching errors in the face of complicated roadways. The article aims at the shortcoming of general map matching, and designs a forecast map matching model based on roadway topology and geometry characteristic. According to the road to roadway topology and geometry characteristic, map matching model divide into three parts, those are initial map matching process (IMP), map matching process at junction (MPJ) and subsequent map matching process (SMP). What´s more, the model is validated by practical running tests. The average correct rate is about 96.1% which shows that the practicability and stability of the device are good. Forecast map matching model brings new idea in order to resolve the difficult problem of matching on complicated roadways.