改进的D-S证据推理地图匹配算法

Improved D-S Evidence Reasoning Map Matching Algorithm

随着中国的高速发展,城市道路拓扑结构越来越复杂,一些常用的地图匹配算法难以适应当今各种复杂的道路类型,尤其是道路交叉口.针对道路交叉口易出错、稳定性差等问题,文中提出了一种改进的D-S证据推理地图匹配算法.该算法对原有算法加以改进,简化证据公式,重新确定基本概率分配函数,并引人第三个证据一历史匹配程度,利用分布式融合模型对证据加以融合,将距离和角度两个基本证据融合后的结果与历史匹配程度证据再加以融合,比较二次融合后的基本概率分配函数值来确定最佳匹配路段.对改进算法执行模拟测试,与其他三种算法相比的结果表明,在错综相连的道路交叉口区域内改进算法拥有更高的匹配准确率且单点匹配耗时更少.

With the rapid development of China,urban road topology is becoming more and more complex,some commonly used map matching algorithms are difficult to adapt to today's complex road types,especially road intersections.It is easy to make mistakes and poor stability for road intersections,an improved D-S evidence reasoning map matching algorithm is proposed.The algorithm improves the original algorithm,simplify the evidence formula,redefine the basic probability assignment function,and introduce a third piece of evidence-the degree of historical matching,integrate evidence using a distributed fusion model,combine the results of the two basic evidences of distance and angle with the evidence of historical matching degree,the basic probability distribution function values after the second fusion are compared to determine the best matching road segment.Perform simulation tests on improved algorithms,the results compared with the other three algorithms show that improved algorithms have higher matching accuracy and less time-consuming matching in the inaccurate intersections.