考虑下肢运动学约束的LKPDR/INS行人导航算法  

LKPDR/INS pedestrian navigation algorithm considering the constraints of lower limb kinematics

作　　者：徐向波[1] 高森宇 朱亚辉 施方艳 张亚楠[1] XU Xiangbo;GAO Senyu;ZHU Yahui;SHI Fangyan;ZHANG Yanan(College of Engineering,Beijing Forestry University,Beijing 100083,China)

机构地区：[1]北京林业大学工学院,北京100083

出　　处：《中国惯性技术学报》2025年第3期267-272,共6页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(32371868)。

摘　　要：针对行人航位推算(PDR)中步长估计模型的准确度受统计结果影响的问题,提出了一种考虑下肢运动学约束的PDR算法(LKPDR)。根据下肢运动学特点,将两个惯性测量单元(IMU)分别固定在行人同侧的大腿和小腿上。通过分析两个传感器相对位置的约束关系,基于角速度及运动学方程进行运动学建模。基于LKPDR和惯性导航系统(INS)设计误差状态卡尔曼滤波器,并在状态向量中引入校正因子,用于补偿导航误差。经过不同场景的行人导航实验验证,所提算法的平均定位相对误差为0.98%~1.74%,较线性步长PDR、非线性步长PDR、零速校正和LKPDR算法提高了11.63%~55.35%,具有更高的精度和环境鲁棒性,且无需增设其他传感器。Aiming at the problem that the accuracy of step length estimation model in pedestrian dead reckoning(PDR) is affected by statistical results,a PDR algorithm considering the constraints of lower limb kinematics(LKPDR) is proposed.According to the kinematic characteristics of the lower limbs,dual inertial measurement units(IMU) are fixed on the thigh and calf of the same side of the pedestrian.By analyzing the constraints on the relative position of the two sensors,the kinematic modeling is performed based on angular velocity and kinematic equations.The error state Kalman filter is designed based on LKPDR and inertial navigation system(INS),and the correction factor is introduced into the state vector to compensate the navigation error.The experimental results of pedestrian navigation in different scenarios show that the average positioning relative error of the proposed algorithm is 0.98% to 1.74%,which is 11.63% to 55.35% higher than that of linear step PDR,nonlinear step PDR,zero velocity update and LKPDR algorithms.The proposed algorithm has higher accuracy and environmental robustness,and no need to add other sensors.

关 键 词：行人航位推算 行人惯性导航 下肢运动学 误差状态卡尔曼滤波 

分 类 号：TN96[电子电信—信号与信息处理]