交叉路段的自适应投影地图匹配算法及应用

An Adaptive Projection Map Matching Algorithm for Intersecting Roads and Its Application

针对目前已有的地图匹配算法在城市交叉路段背景下进行地图匹配时出现的容易波动、空匹配、错误匹配、匹配准确率降低等问题,提出一种交叉路段背景下的自适应投影地图匹配算法。首先利用设置距离最大值法对异常的定位信息进行剔除,并且根据前两个时刻的连续车辆定位点采用插值算法补全缺失的定位数据。基于从开放街图OSM官网下载的城市地图信息,通过与高德地图进行比对,补全开放街道地图OSM所缺失的道路信息,为了提高候选道路查询效率,缩短匹配时间,对电子地图生成网格索引及利用误差圆的计算确定候选道路集合。其次,分别构造车辆投影距离和方向角的概率函数,并且自适应调整各自的权重系数。最后融合车辆投影距离和方向两个影响因素,结合相应的自适应调整权重系数计算候选道路的概率,确定车辆所行驶的实际道路,提高路网匹配的准确率。采用常州市1 800多辆出租车车辆实时行驶数据对此算法进行了试验验证及仿真比较。结果表明:与其他现有3种匹配算法相比,尤其在交叉路段背景下,提出的自适应投影地图匹配算法的准确率可提高4%左右,单点匹配时间缩短1.5 ms左右,大大提高了路网匹配的准确度,也缩短了单点匹配时间,提高了地图匹配算法性能。

In view of the problems of easy volatility, empty matching, wrong matching and low matching accuracy by performing the existing map matching algorithm on the background of urban intersecting roads, an adaptive projection map matching algorithm on this background is proposed. First, the abnormal locating information is eliminated by using the method of setting distance threshold, and the missing locating data are completed by using the interpolation algorithm according to the continuous vehicle locating points of the first 2 moments. Based on the city map information downloaded from the official website of the Open Street Map(OSM) and comparing with the AutoNavi map, the road information missed from the OSM is completed. In order to improve the query efficiency of candidate roads and shorten the matching time, a grid index is generated for the electronic map and the calculation of the error circle is used to determine the set of candidate roads. Second, the probability functions of the vehicle projection distance and direction angle are constructed respectively, and the respective weight coefficients are adjusted adaptively. Finally, considering both projection distance and direction, the probabilities of the candidate roads are calculated with the respective weight coefficients to determine the actual road and improve the matching accuracy or road network. The test verification and simulation comparison of the algorithm are conducted by using the real-time driving data of more than 1 800 taxis in Changzhou City. The result shows that compared with the other 3 existing matching algorithms, especially on the background of intersecting roads, the accuracy of the proposed adaptive projection map matching algorithm can be increased by about 4%,and the single-point matching time can be shortened by about 1.5 ms, which greatly improves the road network matching accuracy, shortens the single-point matching time, improves the performance of the map matching algorithm.