基于移动相关的最小二乘地图匹配新算法  

Novel map-matching algorithm based on moving-related least squares

作　　者：龚柏春[1,2] 罗建军[1,2] 李岁劳[3] 李欣[4] 

机构地区：[1]西北工业大学航天学院,西安710072 [2]航天飞行动力学技术重点实验室,西安710072 [3]西北工业大学自动化学院,西安710072 [4]海军装备研究院,上海200436

出　　处：《中国惯性技术学报》2012年第4期435-439,共5页Journal of Chinese Inertial Technology

基　　金：航天飞行动力学技术重点实验室开放基金(2012afdl037)资助项目;国家自然科学基金项目(61004124)

摘　　要：为了提高军用车辆导航数字地图匹配的精度,针对通常匹配方法易出现误匹配和无法消除沿道路方向纵向误差的问题,提出了一种基于移动相关的最小二乘地图匹配新方法。该方法充分利用路网和历史轨迹信息,以多分辨率移动相关性进行匹配路段的搜索和定位,以最小二乘法进行SINS轨迹到待配路段的曲线整体投影,并通过视觉和数学的方法消除车辆偏离道路中心线的偏差影响。跑车实验表明,匹配后定位圆误差极限在1 m以内。该算法简单有效,且能满足高精度的要求,可以为组合导航提供良好辅助信息。To improve the digital map-matching accuracy of military vehicle navigation and solve the problem that common methods are prone to false matches and cannot eliminate the longitudinal error along the path, a novel least-squares map-matching scheme based on mobile-related is presented. By making full use of the historical trajectory and network information, the method uses multi-resolution mobile relationship to search and locate matched sections, uses least-squares method to do the overall projection from SINS curve to the road section to be matched, and eliminates the deviation of the vehicle off the road centerline with vision or math approach. Experimental results show that the positioning circular error limit after matching is less than 1 m. This algorithm is simple and effective, and can meet the high-precision requirements. It can provide good auxiliary information for the integrated navigation.

关 键 词：组合导航 地图匹配 移动相关性 最小二乘法 

分 类 号：U666.1[交通运输工程—船舶及航道工程]