改进的小波阈值去噪方法在MEMS陀螺上的应用

Application of Improved Wavelet Threshold Denoising Method for MEMS Gyroscop

针对MEMS陀螺由于结构瑕疵和信号处理技术的不足而导致存在较大的随机误差,采用改进的小波阈值去噪方法以提高MEMS陀螺测量精度。通过对软阈值函数和硬阈值函数的缺陷进行分析,设计了改进的阈值函数。改进的阈值函数相对于硬阈值函数具有连续性的优势,相对于软阈值函数可以减小常值误差。根据不同阈值去噪方案在陀螺静态数据上的去噪效果,确定了最终的阈值去噪方案。将此方案应用在MEMS陀螺的动静态数据中,实验结果表明:MEMS陀螺输出信号的标准差减小了一个数量级,平均值更加接近真实值,所提出的采用改进的小波阀值去噪方法取得了很好的去噪效果。

Due to structure defects and limited signal processing technology, MEMS gyro hasthere exists great random error in MEMS gyro. To increase measurement accuracy of MEMS gyro, wavelet threshold denoising method is applied. Firstly, the component of random error in gyro output is analyzed. Then shortcomings of hard threshold and soft threshold function are found through contrast. To solve these shortcomings, improved threshold function is constructed, and through derivation, it can be seen that hard threshold and soft threshold is special case of improved threshold. At last, wavelet threshold denoising scheme is designed and applied in static and dynamic data of MEMS. The result indicates that stability of gyro output increases a degree of magnitude, and the average value is more close to the true value. The proposed the improved wavelet threshold denoising method has good denoising effect.