基于地标点匹配的高精度室内定位算法
High-precision indoor positioning algorithm based on landmark matching
周凌柯 1鲜委 1龚文龙 1李胜1
作者信息
- 1. 南京理工大学 自动化学院,南京 210094
- 折叠
摘要
针对基于惯导的行人室内定位算法定位误差大的问题,提出了一种基于地标点匹配的行人高精度三维室内定位算法.该算法仅使用惯性测量单元(IMU)这一单一传感器,包含速度约束、航向角约束、水平位置约束和高度约束四个模块.水平位置约束模块可根据航向角变化实时检测并建立地标点,并设计了一种地标点匹配度函数,提高了地标匹配的准确度.多楼层行走实验表明所提算法能够提升行人室内定位精度,其终点误差为 0.3421 m,相较于ZUPT算法和ZUPT+HDE算法分别减小了82.6%和 68.3%,具有一定的工程应用价值.
Abstract
Aiming at the problem of large positioning error of inertial navigation system based pedestrian indoor positioning algorithm,a high-precision three-dimensional indoor pedestrian location algorithm based on landmark matching is proposed.The algorithm only uses the inertial measurement unit(IMU)as a single sensor and includes four modules of speed constraint,heading constraint,horizontal position constraint and height constraint.The horizontal position constraint module can detect and establish landmark points in real-time according to the changes of heading angle,and a landmark matching degree function is designed to improve the accuracy of landmark matching.The multi-floor walking experiment shows that the proposed algorithm can improve the accuracy of pedestrian indoor positioning.Compared to the ZUPT algorithm and the ZUPT+HDE algorithm,the terminal point error of the proposed algorithm is only 0.3421 m,which is reduced by 82.6%and 68.3%respectively,and has certain engineering application value.
关键词
室内定位/卡尔曼滤波/航向角约束/地标点Key words
indoor positioning/Kalman filtering/heading constraint/landmark引用本文复制引用
基金项目
高新工程重大专项(5140501B0203)
出版年
2024
NETL
NSTL
万方数据