用于三维姿态估计的双矢量并行卡尔曼滤波  

Dual-vector parallel Kalman filter for three-dimensional attitude estimation

作　　者：李翔[1] 张鹏 唐妍梅[2] Li Xiang;Zhang Peng;Tang Yanmei(School of Electronic Engineering and Automation,Guilin University of Electronic Technology,Guilin 541004,China;College of Physics and Technology,Guangxi Normal University,Guilin 541004,China)

机构地区：[1]桂林电子科技大学电子工程与自动化学院,桂林541004 [2]广西师范大学物理科学与技术学院,桂林541004

出　　处：《国外电子测量技术》2022年第6期60-64,共5页Foreign Electronic Measurement Technology

基　　金：国家自然科学基金(41761087,61603107)项目资助。

摘　　要：基于磁强计、加速度计和陀螺仪的微型航姿参考系统(AHRS)通常采用卡尔曼滤波算法实现传感器间的数据融合。然而,卡尔曼滤波包含的矩阵运算导致其计算复杂度较高,不利于在处理能力有限的单片机上应用。提出一种以地磁矢量和重力矢量为状态变量的双矢量卡尔曼滤波算法,极大地简化了量测模型,有助于降低计算量和避免线性化误差。在AHRS上进行了实验验证,结果表明所提出的双矢量卡尔曼滤波算法在单片机上的运行耗时可比已有的基于四元数的滤波算法减少60%以上,且姿态估计精度优于已有算法。所提出的算法可实现三维姿态的快速、准确估计,改善AHRS姿态输出的实时性。Kalman filter is commonly used in miniature attitude and heading reference system(AHRS)that based on magnetometer,accelerometer and gyroscope to implement multi-sensor data fusion.But the matrix operations in Kalman filter lead to considerable computational complexity,and thus Kalman filter is not fit for microcontrollers with limited processing power.A dual-vector Kalman filter(DVKF)is proposed,which takes the geomagnetic and gravity vectors as its state variables and has extremely simplified measurement model.This algorithm can reduce computational burden and avoid linearization error.Experiment results on AHRS show that the proposed DVKF algorithm can save more than 60%of execution time on microcontroller compared to the existing quaternion-based filtering algorithms,while having higher accuracy in attitude estimation.The proposed algorithm can achieve fast and accurate estimation of three-dimensional attitude,and can help to enhance the timeliness of attitude information from AHRS.

关 键 词：姿态估计 磁强计 加速度计 陀螺仪 卡尔曼滤波 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]