润滑油磨粒超声回波信号的双树复小波自适应降噪

针对润滑油磨粒含有强噪声的回波信号的问题,采用基于双树复小波变换(DT-CWT)的自适应降噪方法,从而提取清晰的磨粒回波信号. 该方法结合奇异谱分析(SSA)和小波熵理论,分别对双树复小波变换后的近似部分和细节部分进行分析. 奇异谱分析去除了近似部分包含的噪声,同时,小波熵理论能够自适应选取不同分解层上的阈值,实现了细节部分系数的自适应选择. 仿真表明,对于润滑油磨粒超声回波信号的双树复小波自适应降噪,输出信号信噪比(SNR)高、均方根误差(RMSE)小、相似系数(NCC)大,算法运算时间能够满足在线检测要求. 实验分析表明,该方法降低了信号中的噪声,还原了准确的波形特征. 

The Adaptive De-Noising Based on Dual-Tree Complex Wavelet Transform for Ultrasonic Echo Signal of Wear Debris in Lubricant Oil

In order to reduce the noise containing in ultrasonic echo signal of wear debris in lubrication, an adaptive method was proposed based on dual-tree complex wavelet transform (DT-CWT) to extract clear and accurate echo signals. Combining singular spectrum analysis (SSA) and wavelet entropy theory, the approximate and detail section of dual-tree complex wavelet transform were analyzed respectively. Singular spectrum analysis was used to remove the noise contained in the approximate section, and the wavelet entropy theory was applied to select thresholds adaptively in different decomposition levels, to achieve the adaptive choosing of detail coefficients. Simulation results show that after the adaptive de-noising based on dual-tree complex wavelet transform, the output signals have higher signal-to-noise ratio (SNR), smaller root mean square error (RMSE), higher normalized correlation coefficient (NCC) and the runtime of algorithm meets the requirement of online detection application. The experimental results show that this method can efficiently reduce the noise of ultrasonic echo signal and restore the accurate wave shape features.