基于CEEMD阈值和相关系数原理的MEMS陀螺信号去噪方法

针对经验模态分解（EMD）方法易产生模态混叠问题，而集成经验模态分解（EEMD）方法又存在重构误差较大的缺陷，提出了一种基于完备集成经验模态分解（CEEMD）阈值滤波和相关系数原理的MEMS陀螺信号去噪方法。首先通过CEEMD方法对陀螺信号进行有效完备的分解，并利用相关系数原理合理确定分解后噪声分量与有效分量的界限。在此基础上，通过借鉴小波阈值处理方式和EMD阈值设置方法，对信号进行阈值滤波去噪。对仿真信号和实际MEMS陀螺信号的研究结果表明，CEEMD阈值去噪方法的去噪效果要优于CEEMD、EEMD、EMD强制去噪方法和小波分析方法。这也充分体现了其在MEMS陀螺信号去噪应用中的可行性和有效性。

Signal De-noising Method for MEMS Gyroscope Based on CEEMD Threshold and Correlation Coefficient Principle

Empirical mode decomposition (EMD) is prone to cause modal aliasing problem, while ensemble empirical mode decomposition (EEMD) method has the shortcomings of larger reconstruction error. In the paper, a MEMS gyroscope signal de-noising method based on complete ensemble empirical mode decomposition (CEEMD) threshold filtering and correlation coefficient principle is proposed. First, the gyroscopic signal is decomposed effectively and completely by CEEMD method, and the boundary between the noise component and the effective component after decomposition is reasonably determined by the correlation coefficient principle. On this basis, by using the method of wavelet threshold processing and EMD threshold setting method, the signal is de-noised by threshold filtering. The research results of simulated signal and real MEMS gyroscope signal show that the de-noising effect of CEEMD threshold de-noising method is better than that of CEEMD, EEMD, EMD forced de-noising method and wavelet analysis method. It also fully embodies the feasibility and effectiveness of MEMS gyroscope signal de-noising application.