碩士 / 義守大學 / 電子工程學系碩士班 / 97 / This research consists of the eigenvalue and recognition system of the vessel acoustic sound, includes the digital signal processing technology. The sound signal of shipping was processed by the transformed frequency spectrum and the multivariate analysis. Then, we simulated and applied the back-propagation neural network to identify the sound.
While dealing with the digital signal processing, we applied the Fourier transform and the wavelet transform to transform the signal of the vessel’s sound into the frequency spectrum. Then, we calculated the eigenvalue from the frequency spectrum by using the multivariate analysis. The sound of the vessel includes the fishing boat, the merchant ship, ship 1, and ship 2. We choose 50 eigenvalues as representative each boat, and lead the eigenvalues in the back-propagation neural network to go on eigenvalue training of boats.
After the back-propagation neural network system experimenting (train are 100 and test are 40), it shows that the identification of the system is working and can solve the problem the identification of the vessel acoustic sound.
Identifer | oai:union.ndltd.org:TW/097ISU05428001 |
Date | January 2009 |
Creators | Wu-Song Yang, 楊武松 |
Contributors | Ke-Nung Huang, Sun-Lon Jen, 黃克穠, 任善隆 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 78 |
Page generated in 0.0018 seconds