|
|
 |
| |
|
|
|
Damage Recognition of Gearbox and Blade for Wind Generation System Using Frequency Spectrum Analysis |
|
|
|
PP: 213-218 |
|
|
|
Author(s) |
|
|
|
Po-Hung Chen,
Hung-Cheng Chen,
Chun-Yao Lee,
|
|
|
|
Abstract |
|
|
| This paper presents a novel approach based on the frequency spectrum analysis for the gearbox and blade damage
recognition of wind generation system. The proposed approach utilizes a back-propagation neural network (BPNN) to extend the
end of the frequency spectrum to overcome the end effect problem of the Hilbert-Huang transform (HHT), which consists of empirical
mode decomposition (EMD) and Hilbert transform (HT). The extension of the two ends obtained by the proposed approach forms a
new frequency spectrum to improve the end effect problem which can distort Hilbert spectrum. The proposed approach is successfully
applied to analyze the generator currents of a wind generation system. Simulated results reveal the new frequency spectrum obtained
by the proposed approach can substantially improve the recognition accuracy of gearbox and blade damage. |
|
|
|
|
|
|
|
