目前车联网中基于D-S证据理论的地图匹配算法具有一定的局限性,匹配精度也无法满足车联网技术的需求。本文对车辆的可达性信息进行考察,作为新的证据与传统算法证据融合后得到的结果进行D-S证据的二次融合,并针对城市环境中不同道路拓扑结构,对传统算法中车辆的位置信息和行驶方向信息的可靠性参数进行仿真训练,得出更为精确的数值以供使用。通过仿真试验表明,改进后的算法的匹配精度和稳定性得到了提高,可以更好地适用于城市复杂路网中地图匹配的问题。
The map matching algorithm based on D-S evidence theory of vehicle-borne navigation system have some limitations and its matching accuracy can not meet the need of vehicle networking technology. This study puts reachability evidence as the third evidence fusion with the integration result of the traditional algorithm. And by simulation training for different road network topology, more accurate position and angle reliability parameters were obtained for practical application. Results of simulation show that this map matching algorithm increases in accuracy and stability, especial suited for complex road network in cities.