A rotating machinery fault sound source localization method based on adaptive modal decomposition and beam-forming

Aiming at the problems of weak fault signals of rotating machinery that are easily submerged by strong background noise, resulting in low acoustic localization accuracy and difficulty in automation, this paper proposes a fault localization method based on adaptive modal decomposition and high-resolution beam-forming. Ensemble empirical mode decomposition is adopted to reconstruct the modal components containing main fault information, effectively suppressing modal aliasing and background noise. Then, the reconstructed pure signals are input into the CLEAN algorithm based on source coherence to achieve high-precision and high-resolution localization of fault sound sources. Experimental results on the bearing fault platform show that compared with traditional algorithms such as DAS and DAMAS, the proposed method exhibits excellent performance under various rotational speeds and fault types, with significantly reduced localization errors and more concentrated energy in the acoustic images, verifying the effectiveness and robustness of the method in fault diagnosis in complex industrial environments.