Application of Wavelet Packet Analysis and Improved LSSVM on Rotating Machinery Fault Diagnosis

For enhancing fault diagnosis precision, the wavelet packet analysis and least squares support vector machine are combined effectively. First, the signals are decomposed in arbitrary minute frequency bands by use of wavelet packet analysis technique. Doing energy calculation in these frequency bands...

Full description

Saved in:
Bibliographic Details
Main Authors: Lingling Zhao, Kuihe Yang
Format: Conference Proceeding
Language:English
Subjects:
Artificial neural networks
Fault diagnosis
KKT conditions
Least squares support vector machine
Mathematical model
Noise
Support vector machines
Wavelet analysis
Wavelet packet analysis
Wavelet packets
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:For enhancing fault diagnosis precision, the wavelet packet analysis and least squares support vector machine are combined effectively. First, the signals are decomposed in arbitrary minute frequency bands by use of wavelet packet analysis technique. Doing energy calculation in these frequency bands to form eigenvectors is more reasonable. And then a least squares support vector machine fault diagnosis model is presented. When the least squares support vector machine is used in fault diagnosis, the Fibonacci symmetry searching algorithm is simplified and improved. It is presented to choose parameter of kernel function on dynamic, which enhances preciseness rate of diagnosis. In the model, the non-sensitive loss function is replaced by quadratic loss function and the inequality constraints are replaced by equality constraints. The simulation results show the model can effectively diagnose machinery facility faults.
DOI:10.1109/PEITS.2008.107