The Grey Support Vector Regression method in noise source identification
Yang Yang, Xiuqin Wang, Di Zhang · 2011
Noise source identification is essential for making noise reduction strategies. This paper presents an approach to acoustic noise identification by introducing modern spectrum estimation, Grey Support Vector Regression (GSVR). Modern spectrum was used to recognize the main noise source and GSVR was used to do curve fitting to recognize the similarity among different curves of power spectrum which made the result more precise. The ranking of the noise sources was obtained on the basis of their individual contribution to the overall noise.