论文部分内容阅读
针对MEMS水听器接收水下信号时混入的噪声,提出一种CEEMDAN和小波包的联合去噪算法.方法是利用CEEMDAN将含噪信号分解成一系列的IMF分量,IMF由高频到低频排列.由于高频噪声和低频干扰,所以要分别对高频和低频IMF进行处理.根据每层IMF的频谱分析,去掉前几层和后几层IMF分量形成一个带通滤波器,然后将剩余分量进行重构信号.最后,利用小波包对重构信号去噪得去噪信号.通过仿真实验和中北大学汾机实测实验验证了算法的有效性,并在去噪效果和性能指标上较单一CEEMDAN和小波包去噪有明显提升.
Aiming at the noise that MEMS hydrophone receives when underwater signal is mixed, a joint denoising algorithm based on CEEMDAN and wavelet packet is proposed. The method uses CEEMDAN to decompose the noisy signal into a series of IMF components, and the IMF is arranged from high frequency to low frequency. Due to high-frequency noise and low-frequency interference, so the high-frequency and low-frequency IMF separately.According to each layer of the IMF spectrum analysis, remove the first few layers and after several layers of IMF components to form a band-pass filter, and then the remaining components Reconstruct the signal.Finally, wavelet packet is used to denoise the reconstructed signal to get the denoised signal.The validity of the algorithm is verified by the simulation experiment and experimental test of Fen machine at North Central University, and the denoising effect and performance index are better than single CEEMDAN And wavelet packet denoising has significantly improved.