摘要
对于低精度高噪声的传感器组成的低成本姿态测量系统,本文引入U nscen ted K a lm an filtering(UKF)用于姿态确定,设计了有陀螺测量和四元数差分法的无陀螺测量两种UKF滤波器;应用四元数避免了欧拉角法的奇异问题;用高斯-牛顿误差最小法将六维参考向量转化为四元数,作为观测量的一部分,使九维非线性观测方程转化为七维线性方程进行滤波,减少了计算量;应用仿真数据进行算法验证,成功得到姿态估计;对两种算法在低速和高速状态下进行验证,仿真结果表明了该方法的有效性。
For a low-cost attitude measurement system made up of less accurate and high noise sensors, the unscented Kalman filtering （UKF） is used for the determination of the attitude. Two UKF algorithms are designed. One of which incorporates rate measurement and the other relays on differencing quaternion. The quaternion is used to avoid the singularity problem of Eular angles. Gauss-Newton error minimization is used to reduce six-dimensional reference vector to the quaternion as a part of observations for UKF. Nine-dimensional nonlinear observation equation is turned into seven-dimensional linear to reduce the computational requirement. Simulation data are used to test algorithms and obtain attitude estimates. Two algorithms are also tested in low and high rate attitude motion. Simulation examples illustrate the effectiveness of the approach.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2006年第1期37-42,共6页
Journal of Nanjing University of Aeronautics & Astronautics
基金
教育部博士点基金(20030287005)资助项目
关键词
姿态测量
UKF
四元数
高斯-牛顿误差最小法
attitude measurement
unscented Kalman filtering
quaternion
Gauss-Newton error minimization
