一种人员惯性定位的鲁棒多约束地图匹配算法  

Robust multi-constraint map matching algorithm for personal inertial positioning

作　　者：赵辉[1,3] 燕瑞超[2] 朱宁 苏中 张孟超[2] 刘高文 ZHAO Hui;YAN Ruichao;ZHU Ning;SU Zhong;ZHANG Mengchao;LIU Gaowen(Beijing Key Laboratory of High Dynamic Navigation Technology,Beijing Information Science and Technology University,Beijing 100192,China;Systems Engineering Research Institute,China State Shipbuilding Corporation Limited,Beijing 100094,China;School of Automation,Beijing Information Science and Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学高动态导航技术北京市重点实验室,北京100192 [2]中国船舶集团有限公司系统工程研究院,北京100094 [3]北京信息科技大学自动化学院,北京100192

出　　处：《中国惯性技术学报》2023年第11期1102-1112,共11页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(61971048);北京市科技计划(Z221100005222024)。

摘　　要：针对传统的地图匹配算法在分岔口和半封闭场景下易出现匹配失效和精度下降的问题,提出了一种人员惯性定位的鲁棒多约束地图匹配算法。该算法构建了地磁/惯性融合航向估计系统,利用三轴陀螺仪实时校准地磁传感器偏差,采用校准后的地磁数据进行航向解算,可有效提升航向信息获取精度;在固定的时间窗口内引入粒子群轨迹的多重约束,增强了算法的鲁棒性;此外,设计了一种启发式多约束重采样策略,提升了粒子群的自适应能力。最后,通过多组室内和室外长距离行走实验对所提算法性能进行评估。实验结果表明,与地图回溯约束的粒子滤波算法相比,本文算法具有更强的鲁棒性,定位精度可提升2倍以上,定位误差小于总行程的0.5%。To address the problem that traditional map-matching algorithms are prone to matching failure and accuracy degradation in bifurcated and semi-enclosed scenes,a robust multi-constraint map-matching algorithm for personnel inertial positioning is proposed.The algorithm constructs a geomagnetic/inertial fusion heading estimation system to calibrate the external geomagnetic interference in real time using a tri-axis gyroscope.The calibrated geomagnetic data are adopted to solve for heading,which could improve heading acquisition accuracy.Multiple constraints are imposed on the particle swarm trajectories within a fixed time window.This effectively increases the robustness of the algorithm.In addition,a heuristic multi-constraint resampling strategy is designed to enhance the adaptability of the particle swarm.Finally,several sets of indoor and outdoor long-distance walking experiments are conducted to evaluate the performance of the proposed algorithm.The results showed that,compared to particle filter algorithm for map backtracking constraints,the proposed algorithm is more robust,the localization accuracy could be improved by more than 2 times,and the localization error is less than 0.5%of total travel distance.

关 键 词：人员定位 惯性测量 地磁信息 粒子滤波 地图匹配 

分 类 号：U666.1[交通运输工程—船舶及航道工程]