A positioning algorithm of autonomous car based on map-matching and environmental perception

Autonomous car is an important tool for transportation and military in the future, and its precise positioning is the basis of autonomous navigation. Most positioning algorithms based on map-matching for autonomous car make little use of environmental perception information. To solve this problem, a positioning algorithm is proposed in this paper, which is based on map-matching and environmental perception for autonomous car. The algorithm includes macroscopic road matching and microscopic precise positioning. As for macroscopic road matching, the algorithm makes use of computational geometry to match the position of autonomous car to the corresponding road, based on GPS point and map information of the road network. As for microscopic precise positioning, the algorithm makes use of the environmental perception, which is detected by the autonomous car to make precise positioning. Macroscopic road matching provides matching road to microscopic precise positioning, and microscopic precise positioning eliminates gross error produced in macroscopic road matching. Through real car tests, the algorithm can match map quickly, improving the positioning precision with strong real-time.