采用四元数无迹卡尔曼滤波的低成本车载姿态航向系统  

Low cost vehicle attitude heading reference system based on quaternion and Kalman filter

作　　者：刘畅 丛丽[1] 秦红磊[1] Liu Chang;Cong Li;Qin Honglei(Electronic and Information Engineering,Beihang University,Beijing 100191,China)

机构地区：[1]北京航空航天大学电子信息工程学院

出　　处：《电子测量技术》2018年第11期40-44,共5页Electronic Measurement Technology

摘　　要：针对低成本车载定姿系统使用惯性导航计算姿态角存在误差累积的问题,采用三轴陀螺仪、三轴加速度计和磁力计实现了姿态航向系统,有效抑制了误差对陀螺姿态的影响。该系统基于四元数的微分方程,通过四阶龙格库塔法进行时间更新,采用无迹卡尔曼滤波融合算法,引入加速度计和磁力计计算的角度进行滤波更新,提高了低成本车载姿态航向系统的精度。实验测试结果表明,基于四元数无迹卡尔曼滤波的低成本车载航向系统在静止状态下航向角、俯仰角、横滚角最大误差分别可以控制在0.021°,0.038°,0.015°以内,动态情况下,航向角、俯仰角、横滚角的均值误差分别为0.129°,0.729°,0.303°。Low cost vehicle attitude heading reference system is realized to solve the problem of the accumulation of error.The system consists of three-axis gyroscope,three-axis accelerometer and magnetic compass.It is based on the quaternion differential equation and updated by the fourth-order Runge-Kulta method.The unsigned Kalman filter fusion algorithm is adopted by using the angle calculated by three-axis accelerometer and magnetic compass.It improves the accuracy of low-cost vehicle attitude heading system.The experimental results show that,under stationary state,the maximum error of yaw,pitch and roll are0.021°、0.038°、0.015°respectively.Under dynamic situation,the error average of yaw,pitch and roll are 0.129°、0.729°、0.303°respectively.

关 键 词：低成本 四阶龙格库塔法 四元数微分方程 航向姿态参考系统 无迹卡尔曼滤波 

分 类 号：TN967.2[电子电信—信号与信息处理]