Research Of Map Matching Technology In Complex Road Network In Vehicle Navigation System

Currently vehicle navigation system has been applied widely, and generally map matching technology is used to solve the problem that the driving tracks are not on the road caused by satellite positioning and mapping errors, etc. Map matching is a process of correcting the received locator data with the aid of electronic map real- time. Map matching algorithm can reduce the influence of the location error on the positioning and improve the positioning accuracy of the system. The existing map matching algorithms have high precision and good results, which obtain a better application in most vehicle navigation systems. However, when the satellite signal is sheltered, vehicles can’t effectively distinguish the case of on or under the overpass on the complex overpass if navigation system can’t position effectively under the tunnels and overpasses. At the same time, for the military vehicle navigation systems with complex applications, they not only include the navigation applications, but also include the complex map data processing and a variety of business data processing, etc. More complex matching calculation may have a pause when switching matching area, thus putting forward higher requirements on the matching algorithm. For the actual application requirements, this paper studies the implementation process of map matching algorithm deeply, and designs a map matching algorithm with strong real-time and good matching results for complex road network. The main works of the paper are as follows:(1) Aiming at the problem of matching pause or delay caused by long time reorganization of road network data when matching calculation, adopting the strategy that storage space is exchanged with response time, the access method of road network data based on double buffering mechanism is realized in multitasking environment based on priority scheduling. First of all, the method applies for two pieces of cache space at system startup, and adopts a node-section associated based matching road network model to organize network data in the front buffer. And then, primary and auxiliary tasks with high and low priority are established. Among them, the main task with high priority does matching and calculating real-time based on node-section associated model in the front buffer, and the auxiliary task pretreats the regional road network data read by the back buffer to reduce the number of matching search effectively. At the same time, the node-section associated connection of a new position is created in advance to avoid the problems caused by matching switch of the new position effectively. The experimental results show that the double buffering strategy achieves the smooth running of map matching, thus improving the access efficiency of road network data and guaranteeing the real-time matching of vehicles.(2) Aiming at the problem that the real-time of map matching is influenced by excessive number of road shape points and searches when in road network search in an embedded environment, a preprocessing method of vector data compression with adaptive tolerance based on matching error is studied. This method adds the vector data compression module with a feedback loop in the map matching process, and applies the improved Douglas-Peucker algorithm with adaptive tolerance for regional road network data preprocessing. Respectively taking map accuracy, road spacing and the maximum likelihood estimate of the matching error as input parameters, the improved algorithm takes the weighted average as the precision tolerance to compress road network data after giving appropriate weights. The experimental results show that the points compressed reduce significantly compared with the points uncompressed, which describes that the road network pretreatment can effectively reduce the data capacity and speed up the search, thus improving the real-time of map matching.(3) Aiming at the problem that the matching precision of the two-dimensional map matching algorithms is relatively low towards the three-dimensional sections such as overpasses and tunnels, the improved D-S evidence theory map matching algorithm based on elevation information is designed. First of all, according to the presence of historical location information, the algorithm adopts different search modes to determine the candidate road set. Secondly, through the custom elevation function, improved algorithm adds an elevation factor to the joint support function fused by the evidences of location information and direction information to introduce the role of elevation information. Then, the matching section is determined according to the results of the improved joint support function. In the end, the orthogonal projection method and dead reckoning method are adopted to correct the positioning results for valid locator data and missing locator data. The experimental results show that the improved algorithm can improve the matching accuracy of the overpass areas effectively compared with the traditional method of D-S evidence theory. At the same time, it also can improve the processing performance of map matching algorithm.(4) Component-based map matching module based on register mechanism is designed and implemented in × × navigation and positioning system developed by research group. After testing and verification, the component can satisfy the application demand of the system.