A novel approach for accurate identification of complex noise sources

Bing Li,Diange Yang,Ziteng Wang,Sifa Zheng,Xiaomin Lian

2009-01-01

Abstract:Beam-forming technique has been widely used in noise control applications due to its high computational efficiency. However, sometimes non-real ghost noise sources are created in the identification results because of the side lobe effects. The wave superposition algorithm and acoustic holography method have been known as a usual approach for noise identification but have more complex computation. Especially when applied to reduce the side lobe effects, more microphones need to be installed, which will lead to an increased measuring cost. In our paper, we present a new approach for noise source identification by combining beam-forming, wave superposition algorithm, and acoustic holography altogether. Firstly, we used beam-forming to pre-estimate the positions of the noise sources. Then wave superposition algorithm was applied to estimate the magnitude of these noises. Finally, with the results obtained from the previous two steps, the acoustic holography was applied to reconstruct the source sound field. Our emulational experiments proved that, with same amount of microphones, our approach have better noise identification results and higher resolution compared to beamforming method respectively, suggesting our method's great potential in noise control by detecting and identifying the noises accurately.