车载GPS导航系统中GPS定位与道路匹配方法研究

Research on GPS Positioning Matching with Road in Mobile Navigating System

对车载GPS导航系统中GPS定位与道路的匹配方法进行了深入的研究 ,利用已有的道路网数据 ,建立了道路网络拓扑关系 ,在此基础上提出了基于网络拓扑关系和电子罗盘方位信息的快速道路匹配算法。通过车载试验证明 ,该算法能够实现GPS定位与道路实时、可靠、高精度的匹配。

At present,the positioning precision of single GPS can only reach 100 meters.As the limited cost,it is impossible to process the signal of GPS by real_time difference to obtain high positioning precision.Therefore,in car navigating system,single GPS receiver is generally used to get the position of car in real_time.Because of the effect of American SA technology,the positioning precision of single GPS has only 100 meters.In this case,if we can not adopt suitable method to process the positioning information in real_time,we can not get better result in car navigating system.For example,when the car is running along the bridge,because of positioning error the car maybe positioned in the river. As the car always runs on the road in normal situation and the database of road net has been established in advance before navigating.So we can use above condi_tions as restricted conditions. Before starting to navigate we can build topology based on the database of road net and get the initial position of the car on the road.At the same time,because the car is on the road,the error in vertical road direction is eliminated automatically.As the error of along_road_direction is concerned,it is not important for navigating.On the other hand,it maybe be eliminated according to the apparent mark along the road (e.g.cross_road,megastructure nearby road) by hand.When navigating we can correct the signal of GPS according to the current road that the car is running.It is critical to process the signal of GPS when the car reaches cross_road.Under this condition,we can quickly find several roads that the car will run according to road topology. As the positioning error is 100 meters,we can not decide immediately where the car will run.Therefore,due to positioning error the car maybe jump from one road to another road.Here we can add another restricted condition by using the data of compass installed on the car.As compass can record the driving direction of the car,when the car come to cross_road nearly we can analyze the acquired data from compass in real_time to decide where the car will go.At the same time,we can calculate every road direction that the car will choose.By road direction matching with the result direction of analyzed compass data,we can obtain which road the car will go.