基于EMD算法的光栅莫尔条纹信号去噪方法研究  

Study of De-noising Based on EMD Algorithm in Grating Moiré Fringe Signal

作　　者：杨华晖 刘福 冯伟利 

机构地区：[1]军械工程学院,河北石家庄050003 [2]军械技术研究所,河北石家庄050003 [3]北京航天测试计量技术研究所,北京100076

出　　处：《计量学报》2016年第5期467-471,共5页Acta Metrologica Sinica

基　　金：军工计量科研项目(J032008A005)

摘　　要：针对计量光栅莫尔条纹信号的质量问题，提出了基于经验模态分解（EMD）算法对非平稳光栅莫尔条纹信号模型的去噪方法。建立非平稳的光栅时变信号模型，利用EMD算法不需要定义滤波器参数的自适应性优点，对添加不同噪声的多组光栅信号模型进行了滤波分析的仿真实验，其信噪比和均方根误差两项指标优于均值滤波、小波阈值去噪方法。对两路正余弦理想信号添加高次谐波分量，通过对比EMD算法抑制高次谐波前后的李萨如图形，验证了该方法在去噪过程中对光栅莫尔条纹信号正弦性误差补偿的良好效果。In view of the quality problem of metrological grating Moire fringe signal, a method based on empirical mode decomposition （EMD） algorithm is presented for non-stationary metrological grating Moire fringe signal de-noising. A non-stationary factual dynamic grating signal model was established. The simulation experiment of filtering analysis was conducted for several grating signal models with different noise by using the EMD algorithm advantages which is not required to define the filter parameters. The two indicators of SNR and RMSE shown that the EMD filtering outperforms the median and wavelet threshold methods. Add the high-order harmonic components to the sine ＆ cosine ideal signals and the ideal compensation results of sine deviation error in metrological grating Moire fringe signal de-noised processing can be verified by comparing before and after Lissajous figures of suppressing the high-order harmonics with EMD algorithm.

关 键 词：计量学 光栅 莫尔条纹信号 经验模态分解算法 去噪 滤波分析 

分 类 号：TB92[一般工业技术—计量学]