组合导航用MEMS陀螺随机误差特性分析的分段Allan方差法

针对Allan方差分析法当陀螺采样率一般比较高,数据量比较大时,单一分组方式很难同时顾及计算效率和精度的问题,提出一种分段Allan方差法。新方法利用“2k”Allan方差在高频部分计算效率高和“k”Allan方差反映低频部分结果更精细的优点,通过新方法与“2k”Allan方差法、“k”Allan方差法在计算耗时、解算陀螺误差参数可靠性等方面的对比实验,结果表明,新方法在计算效率时间和计算精度方面优于单独的“2k”或“k”Allan方差法。

Segmental Allan Variance Method for Random Error Analysis of MEMS Gyro in Integrated Navigation

In Allan variance analysis method,when gyro sampling rate is generally high and data amount is large,it is difficult for single grouping method to take into account calculation efficiency and precision at the same time,a piecework Allan variance method is proposed.The new method makes full use of the advantages of the high computational efficiency of“2k”Allan variance in the high frequency part and“k”Allan variance in reflecting the finer results in the low frequency part.By comparing the new method with the“2k”Allan variance method and the“k”Allan variance method in terms of calculating time consuming and calculating the reliability of gyro error parameters,the results show that the new method is better than the single“2k”or“k”Allan variance method in calculating efficiency time and accuracy.