足部安装MEMS-IMU个人导航系统  

Pedestrian navigation algorithm based on MEMS-IMU

作　　者：王立兵[1] 杨松普[2] 罗巍[2] 皮燕燕[2] WANG Li-bing YANG Song-pu LUO Wei PI Yan-yan(Troop 63961 ofPLA, Beijing 100012, China Tianjin Navigation Instrument Research Institute, Tianjin 300131, China)

机构地区：[1]中国人民解放军63961部队 [2]天津航海仪器研究所

出　　处：《中国惯性技术学报》2016年第4期460-463,共4页Journal of Chinese Inertial Technology

基　　金：国防基础科研重点项目(A0320132002)

摘　　要：为了实现室内外环境下个人自主导航,研究了足部安装的MEMS-IMU个人导航系统。根据人行走时足部具有周期性零速的特征,以加速度计输出矢量和、滑动方差和陀螺仪输出的角速度矢量和为检测量,设计了一种多条件零速检测算法,有效地提高了零速检测的准确性。针对MEMS惯性传感器零漂大、精度低的问题,导航定位算法以传统的捷联解算算法为基础,进行了适应性改进。引入零速修正(ZUPT)技术,设计了以速度信息作为伪量测的Kalman滤波器。在零速阶段对系统速度,姿态,位置误差进行估计,将估计结果反馈以修正导航解算的累积误差。实验结果表明,基于上述导航修正算法可以有效地消除MEMS惯性传感器零漂引起的累积误差,使得多组多种行走路径下系统的定位误差均小于行程的2%。The pedestrian navigation system（PNS） based on foot-mounted MEMS-IMU is studied. According to the gait model, the foot has to be stationary periodically in between steps during walking. To detect the stance phase, a zero velocity detection algorithm is proposed by using the vectorial sum of accelerometer output, the sliding variance, and the vectorial sum of gyro＇s angular rate as measurements. As the position error of SINS diverges with time, the zero velocity updating（ZUPT） technique is implemented, and the attitude error, the velocity error and the position error estimated by the designed Kalman filter are fedback to correct the accumulated errors of SINS. Experiment results show that the designed amendment navigation algorithm can efficiently reduce the drift error of the MEMS inertial sensors, and the PNS position accuracy can reach 2% of travel distance under various walking path situations.

关 键 词：个人导航系统 MEMS 零速修正 零速检测 

分 类 号：U666.1[交通运输工程—船舶及航道工程]