基于多属性融合策略的车载导航地图匹配算法  

Vehicle navigation map matching algorithm based on multiple-attribute integration strategy

作　　者：滕志军 曲兆强[1] 郭素阳 于明 付饶 井梓鉴 

机构地区：[1]东北电力大学信息工程学院,吉林吉林132012 [2]中交天和机械设备制造有限公司,江苏苏州215000 [3]国网吉林供电公司,吉林吉林132001

出　　处：《江苏大学学报（自然科学版）》2018年第1期14-18,25,共6页Journal of Jiangsu University：Natural Science Edition

基　　金：国家自然科学基金资助项目(51277023)

摘　　要：目前直接投影算法常用于嵌入式车载系统,但其时效性以及精确性不能满足日益复杂的城市道路网络.针对传统算法检索复杂道路耗时多等问题,提出了基于多属性融合策略的车载导航地图匹配算法,提高了嵌入式车载地图匹配算法的时效性,算法采用等步长分块策略,建立分块网格索引减少筛选候选区域时间,从而提高算法的时效性;通过引入车辆行驶速度和历史匹配程度改进基于权重的直接投影算法,使用多属性融合策略确定匹配算法权重因子,提高地图匹配的准确性.结果表明:该地图匹配算法能够快速、准确地匹配出行驶路径,单点匹配时间降低至15 ms,复杂道路下,匹配准确率达到90%以上,能较好地适应于城市复杂道路网.The present direct projection algorithm is often used in embedded vehicle system,while the timeliness and accuracy can not meet the increasingly complex urban road network. To solve the timeconsuming problem of complex road in traditional retrieval algorithm,a matching algorithm for vehicle navigation map fusion strategy was proposed based on multi attribute matching algorithm to improve the efficiency of embedded vehicle map. The equal step block strategy was adopted for building block grid index to reduce screening candidate regions and improve the algorithm efficiency. The direct projection algorithm based on weight was improved by the introduction of vehicle speed and history matching degree,and the multiple attribute fusion strategy was used to determine the matching algorithm of weighting factor and improve the map matching accuracy. The experimental results show that the map matching algorithm can quickly and accurately match the driving path,and the single point matching time is reduced to 15 ms with accurate rate more than 90% under complex roads. The proposed algorithm is suitable for the complex city road network.

关 键 词：智能交通 地图匹配 融合策略 车载 匹配程度 

分 类 号：TN967.1[电子电信—信号与信息处理]