Application Of Improved VMD And Wavelet Threshold Denoising Method In Bolt Detection

Abstract

Rock bolts are often used for reinforcement of slopes, tunnels, and mines in transportation, water conservancy, coal mining and other industries. Thus, identifying the quality of rock bolts is important. To address the difficult of obtaining useful information on rock bolts in complex environment, this study proposes a noise reduction method based on the combination of improved variational mode decomposition (VMD) and wavelet threshold denoising (WTD) method. In this method, the number of decomposition modes in improved VMD algorithm is determined by weighted kurtosis values. First, the electromagnetic ultrasonic signal is decomposed into a finite number of intrinsic mode functions (IMFs) through improved VMD. Second, noisy and invalid IMF components are removed according to weighted kurtosis values. Finally, the residual IMF components are used to reconstruct signal and wavelet threshold processing to improve denoising effect. This VMDWTD method is applied in both simulation signal and magnetostrictive guided wave signal. The results showed that the method retained more valid information and had better noise reduction effect than ensemble empirical mode decomposition with WTD and traditional WTD method.