基于奇异摄动的捷联惯导伪圆锥误差补偿算法

摘要

A novel high-accuracy attitude updating algorithm for a strapdown inertial navigation system (SINS) is given. Designed based on the singular perturbation theory, the new algorithm is essentially an improvement on the traditional Savage attitude updating algorithm which has the problem of pseudo-coning error. Pseudo-coning error is a coning-amplified phenomenon which occurs when the frequency of a vehicle' s actual angular rate the and high bandwidth noise frequency overlap each other. When using the new algorithm, the singular perturbation theory is introduced to restrain the pseudo-coning error generated from the attitude updating. Two time scales characteristics of Gyro-sensed attitude angle motion and Gyro non-ideal output signal are analyzed, furthermore the gyro output is divided into fast-changing portion and slowly-changing portion. And the two time scales characterics are used to compensate the pseudo-coning error. In the process of adjusting the parameters of this algorithm, the Pseudo-coning error in an arbitrary frequency band is control. It is shown that, compared to the Savage four-sample algorithm, the presented algorithm will not only greatly increase attitude updating rate, but also has the favorable flexibility and relatively smaller amount of calculation.%提出了一种新的高精度捷联惯导系统(SINS)姿态更新算法——基于奇异摄动理论的算法.该方法是对伪圆锥误差较大的传统Savage姿态更新算法改进,它运用奇异摄动理论抑制载体真实的角运动频率与宽频带噪声的频率混叠在一起时出现的圆锥误差放大现象——伪圆锥误差,同时将陀螺输出分为快变部分和慢变部分,利用陀螺敏感的载体角运动与陀螺非理想输出信号的两时间尺度特性补偿伪圆锥误差.该算法通过可调参数以控制任意频段的伪圆锥误差的影响,实现算法精度的提高.仿真结果表明,此新方法与目前高精度的Savage四子样算法相比不仅能够提高姿态更新的精度和速率,而且具有良好的灵活性和较小的计算量.