Spectrum pattern recognition for five categories of perceived noise produced by small electric motors

スポンサーリンク

概要

• 論文の詳細を見る

• As an alternative to the present listening inspection techniques, the spectrum pattern recognition technique which automatically detects defective parts of an electric motor was studied using small AC motors. In the noise signal analysis, 1/10-octave high resolutive spectra revealed the apparent features distinguishing the five categories of the perceived noise establishing by listening experts. By using the predetermined reference patterns, the recognition technique correctly determined the condition of thirty samples of small AC motors. Thus, the subjective categories agreed well with the error classes confirmed by the physical measurements.

• 社団法人日本音響学会の論文

著者

• Kaneko Tatsuya

  Life Style Research Center Hitachi Ltd.

• Tannaka Yasuaki

  Consumer Products Research Center,Hitachi Ltd.

• Koshikawa Tsuneji

  University of Electro-Communications,Department of Electronics Engineering

• Koshikawa Tsuneji

  Department of Electronics Engineering,University of Electro-Communications

• Tannaka Y

  Hiroshima‐denki Inst. Technology Hiroshima Jpn

• Koshikawa Tsuneji

  Department Of Electronics Engineering University Of Electro-communications

関連論文

• Spectrum pattern recognition for five categories of perceived noise produced by small electric motors
• Correlations between physical measurements and subjective evaluations of audio compact-cassette decks
• Correlations between harmonic distortion, sound field characteristics and reproduced sound quality change in listening tests for loudspeakers

スポンサーリンク