An acoustic fault sample expansion method based on noise disturbance

Publication Type:
Conference Proceeding
Citation:
24th International Congress on Sound and Vibration, ICSV 2017, 2017
Issue Date:
2017-01-01
Full metadata record
Files in This Item:
Filename Description Size
Weina.pdfPublished version388.88 kB
Adobe PDF
Sample expansion is an effective approach to resolve the incomplete sample problem in ship acoustics fault source identification. A novel sample expansion method based on noise disturbance is proposed. Firstly, the expanded samples are generated by adding different signal-tonoise ratio noise to the real acoustics fault samples. Then, the training collection is enhanced by combining the original training samples with the expanded samples. Experimental results on both simulated data and measured data demonstrate that the classifier performance can be improved by the expanded training sample set.
Please use this identifier to cite or link to this item: