Pipeline Leak AE Signal Denoising Based on Improved SSA-<i>K</i>-α Index-VMD-MD

Abstract

To denoise acoustic emission (AE) pipeline signals and improve the accuracy of nondestructive pipeline leak identification, this article proposes a new denoising method based on variable mode composition (VMD). First, to improve the performance of traditional VMD parameters optimization functions, a new fitness function— <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${K}$ </tex-math></inline-formula> - <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula> index—is proposed to ensure the correct setting of VMD parameters from three perspectives: preventing overdecomposition, preventing underdecomposition, and preventing information loss of original signal. Also. the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${K}$ </tex-math></inline-formula> - <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula> index is used as the optimization objective of salp swarm algorithm (SSA) to optimize VMD parameters. Then, the leak signal is decomposed by VMD according to the optimized parameters into several intrinsic mode functions (IMFs). Next, to more reliably divide effective IMFs and noise components, the Mahalanobis distance (MD), improved by the probability distribution, is used to measure the similarity between IMFs and the original signal. Finally, to accurately identify the leak, an SSA optimized support vector machine (SSA-SVM) is designed. The proposed method is verified by the simulation analysis and experimental verification, and the results show that the proposed method is reasonable. After the AE signal was processed using the proposed method, the leakage recognition accuracy increased from 83.33% to 100%, and the accuracy of leak cause identification increased from 60.41% to 83.33%.