Project/Area Number |
12650144
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Single-year Grants |
Section | 一般 |
Research Field |
設計工学・機械要素・トライボロジー
|
Research Institution | Kyushu Institute of Technology |
Principal Investigator |
CHEN Png Kyushu Institute of Technology, Facul;ty of Computer Science and System Engineering, Associate Professor, 情報工学部, 助教授 (50231428)
|
Co-Investigator(Kenkyū-buntansha) |
NIHO Tomoya Kyushu Institute of Technology, Facul;ty of Computer Science and System Engineering, Assistant, 情報工学部, 助手 (60295011)
豊田 利夫 九州工業大学, 情報工学部, 教授 (00227662)
|
Project Period (FY) |
2000 – 2001
|
Project Status |
Completed (Fiscal Year 2001)
|
Budget Amount *help |
¥2,300,000 (Direct Cost: ¥2,300,000)
Fiscal Year 2001: ¥500,000 (Direct Cost: ¥500,000)
Fiscal Year 2000: ¥1,800,000 (Direct Cost: ¥1,800,000)
|
Keywords | condition diagnosis / unsteady operating conditions / Short Time Fourier Transform / Wavelet Analysis / Wigner Distribution / Instantaneous Power Spectrum / Relative Crossing Information / Genetic Programming / 可変運転条件 / 設備診断 / 瞬間スペクトル / 軸受 / 歯車装置 / 可能性理論 |
Research Abstract |
In order to detect fault and identify fault type for plant machinery in unsteady operating conditions, new methods have been proposed in this study as follows : 1. the diagnosis sensitivities of Time-Frequency Analysis methods , such as Short Time Fourier Transform (STFT), the Wavelet Analysis (WA) and the Wigner Distribution (WD), are investigated for the condition diagnosis of machinery in unsteady operation condition. In order to diagnose failures and evaluate the diagnosis sensitivity, the extracting method of feature spectra by the Relative Crossing Information (RCI), the symptom parameters (SP) in frequency domain, the K-L expansion method for synthesizing SPs and the sequential diagnosis method are proposed. In the case of the bearing diagnosis, the diagnosis sensitivity of the Short Time Fourier Transform (STFT) is found highest. It is proved that the methods proposed in this paper are effective by applying them to the bearing diagnosis. 2. The feature of each machine state can be expressed by the Instantaneous Power Spectrum (IPS), and the feature spectra can be extracted by the Relative Crossing Information (RCI). The excellent symptom parameter (GP-SP) for expressing the characteristics of the feature spectra can be automatically generated by Genetic Programming (GP) for condition monitoring of plant machinery. The methods proposed here have been proved by applying them to the practical failure diagnosis of machinery. The new diagnosis method applying methods of forming symptom parameters automatically by Genetic Programming and concluding the most appropriate band width of frequency by Wavelet analysis. And the validity of the method has been verified by applying it to a rolling bearing in unsteady operating conditions.
|