A block bayesian compressive sensing method based on the equivalent source method for near-field acoustic holography

Abstract

The conventional equivalent source method for near-field acoustic holography is an effective noise diagnosis method using microphone array. However, its performance is limited by microphone spacing, so the effect is unsatisfied when the wave number is high. In this paper, to broaden the frequency suitability and improve the performance of sound source reconstruction with low signal-to-noise ratios, a block Bayesian compressive sensing method based on the equivalent source method is proposed. Numerical results show that this proposed method has a good reconstruction performance and makes wideband reconstruction possible. By changing the frequency, location, and signal-to-noise ratio of the sound source, the reconstruction performance of the proposed method can remain stable. Finally, the validity and practicability of the proposed method are verified by experiments.

High precision localisation of the sources of noise of a bent-axis motor

“…Wang et al 19 used compressed sensing to achieve super-resolution reconstruction of a single image. Ming et al 20 proposed a block Bayesian compressive sensing method based on the equivalent source method to improve the performance of sound source reconstruction with low signal-to-noise ratios. Ning et al 21 applied sparse sampling to the microphone array acoustic imaging to realize the noise location of the air compressor.…”