Study of Fault Signal Noise Reduction Based on Improved CEEMDAN-SVD
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
维普
万方数据
针对旋转机械结构复杂及振动信号强耦合所导致的噪声识别和提取特征困难的问题,提出了一种自适应噪声完备集合经验模态分解(Complete Ensemble Empirical Mode Decomposition with Adaptive Noise,CEEMDAN)与改进阈值筛选SVD算法对复杂噪声作用下的旋转机械振动信号进行降噪处理,运用信噪比与短时傅里叶变换(Short Time Fourier Transform,STFT)、连续小波变换(Continuous Wavelet Transform,CWT)分析来验证所用方法的有效性.实验结果表明:结合改进CEEMDAN与改进阈值筛选SVD算法对旋转机械振动信号进行降噪的效果较好,能有效的保留旋转机械振动信号特征信息,为后续利用旋转机械振动信号进行旋转机械的故障诊断奠定基础.
To overcome the challenges posed by the intricate structural features and strong coupling of vibration signals in rotating machinery,we propose an adaptive noise reduction approach named Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN).Additionally,we introduce an enhanced threshold screening Singular Value Decomposition(SVD)algorithm to mitigate noise interference in vibration signals from rotating machinery,particularly in the presence of complex noise sources.The efficacy of our proposed method is assessed using metrics such as signal-to-noise ratio,as well as through Short Time Fourier Transform(STFT)and Continuous Wavelet Transform(CWT)analyses.Experimental results demonstrate that the amalgamation of improved CEEMDAN with the enhanced threshold screening SVD algorithm yields superior noise reduction performance in vibration signals from rotating machinery.Importantly,this approach effectively preserves the intrinsic characteristics of the vibration signals,thereby laying a solid groundwork for subsequent fault diagnosis of rotating machinery based on vibration analysis.
维普
万方数据
