Abstract: In order to solve the problem of map matching in traffic information collection based on wireless positioning，a new map matching algorithm is proposed，which forms three characteristic variables time series data using instantaneous position，direction and length of the moving trajectories，then DTW methods are used to discover the similarities of the trajectories between the vehicle and candidate roads，finally the ordering method of fuzzy preference relations is used to evaluation the general similarity，the trajectory of mobile telephone and the trajectory which most similarity to it will be matched.The results obtained through this approach are promising.

Key words: map matching, wireless position, dynamic time warping, fuzzy preference relations

摘要： 针对基于无线定位的交通信息采集中地图匹配问题，提出一种新的地图匹配算法，算法提取连续多个位置点的瞬时位置、方向和相邻时刻的运动距离构成具有三个特征变量的子时间序列，并用DTW来进行车行轨迹与候选路段的相似性测量，然后利用模糊偏序关系得到相似性的综合评估，将与移动台运动轨迹最相似的路段作为匹配点所在路段。实验结果证明了算法的有效性。

关键词: 地图匹配, 无线定位, 动态时间弯曲, 模糊偏序关系