Spectrum classification for early fault diagnosis of the LP gas pressure regulator based on the Kullback-Leibler kernel

Tsukasa Ishigaki, Tomoyuki Higuchi, Kajiro Watanabe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

The present paper describes a frequency spectrum classification method for fault diagnosis of the LP gas pressure regulator using Support Vector Machines. Conventional diagnosis methods are not efficient because of problems such as significant noise and nonlinearity of the detection mechanism. In order to solve these problems, a machine learning method with the Kullback-Leibler (KL) kernel based on the KL divergence is introduced into spectrum classification. We use the normalized frequency spectrum directly as input with the KL kernel. The proposed method demonstrates a higher accuracy than popular kernels, such as polynomial or Gaussian kernels, or the conventional fault diagnosis method and Gaussian Mixture Model with the KL kernel for the examined problem. The high classification performance is achieved by using an inexpensive sensor system and the machine learning method. This method is widely applicable to other spectrum classification applications without limitation on the generality if the spectrums are normalized.

Original languageEnglish
Title of host publicationProceedings of the 2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing, MLSP 2006
PublisherIEEE Computer Society
Pages453-458
Number of pages6
ISBN (Print)1424406560, 9781424406562
DOIs
Publication statusPublished - 2006 Jan 1
Externally publishedYes
Event2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing, MLSP 2006 - Maynooth, Ireland
Duration: 2006 Sep 62006 Sep 8

Publication series

NameProceedings of the 2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing, MLSP 2006

Other

Other2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing, MLSP 2006
CountryIreland
CityMaynooth
Period06/9/606/9/8

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Signal Processing

Fingerprint Dive into the research topics of 'Spectrum classification for early fault diagnosis of the LP gas pressure regulator based on the Kullback-Leibler kernel'. Together they form a unique fingerprint.

Cite this