基于小波降噪与最小二乘估计的石英挠性加速度计模型辨识

Model Identification of Quartz Flexible Accelerometer Based on Wavelet Denoising and Least-Square Estimation

石英挠性加速度计在惯性导航系统中具有重要作用,其使用精度影响着整个系统的精度。为提高加速度计的使用精度,对其进行了重力场静态翻滚测试,采用融合小波降噪与最小二乘估计的方法实现加速度计模型参数辨识。利用拟合残差标准差和系数估计的不确定度来评价辨识的效果,对重力场静态测试时加速度计两种安装方式下的测量信号进行了实验测试,并通过仿真实验来验证所采用方法的有效性。实验结果表明,所采用的方法获得的加速度计模型系数具有较高精确度。

The quartz flexible accelerometer is very important in inertial navigation systems, and it affects the precision of whole system. To improve its usage accuracy, an identification method is explored with model parameters of aecelerometer. Using the static test of gravitational field, the techniques of wavelet denoising and least-square estimation were combined to implement the aeeelerometer model identification. To evaluate this proposed method, the standard deviation of residuals and the uncertainty of coefficient estimates were employed. The experiments on accelerometer measurement signals under two kinds of installation method in the gravitational field were done. Also, a simulation experiment was applied to verify the effectiveness of proposed method. The results showed that the estimation of accelerometer model coefficients with the proposed method could achieve high accuracy.