Research Article
Enhancing the Capacity of Detecting and Classifying Cavitation Noise Generated from Propeller Using the Convolution Neural Network
@INPROCEEDINGS{10.1007/978-3-030-77424-0_22, author={Hoang Nhat Bach and Duc Van Nguyen and Ha Le Vu}, title={Enhancing the Capacity of Detecting and Classifying Cavitation Noise Generated from Propeller Using the Convolution Neural Network}, proceedings={Industrial Networks and Intelligent Systems. 7th EAI International Conference, INISCOM 2021, Hanoi, Vietnam, April 22-23, 2021, Proceedings}, proceedings_a={INISCOM}, year={2021}, month={5}, keywords={Passive sonar DEMON Convolution neuron network}, doi={10.1007/978-3-030-77424-0_22} }- Hoang Nhat Bach
Duc Van Nguyen
Ha Le Vu
Year: 2021
Enhancing the Capacity of Detecting and Classifying Cavitation Noise Generated from Propeller Using the Convolution Neural Network
INISCOM
Springer
DOI: 10.1007/978-3-030-77424-0_22
Abstract
One of the biggest concerns of underwater research is improving the ability to detect and classify sound sources. The Machine Learning and Deep Learning models often require a very large amount of data, while the data sources of the passive sonar system are limited; therefore, it is very important to pre-process data to improve data quality. This paper proposes a solution to improve the detection and classification of cavitation noise generated from propeller by improving the Detection of Envelope Modulation on Noise (DEMON) algorithm before using a modified Convolution Neural Network. The testing result shows that the accuracy of the modified model reaches around 90%, which is better than the results of existing methods, and it is prospectively developed and applied in practicalities.
