| Project/Area Number |
14350081
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| 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 | Mie University (2003-2004)
Kyushu Institute of Technology (2002) |
|---|
Principal Investigator |
JINYAMA Ho Mie University, Faculty of Bioresources, Professor, 生物資源学部, 教授 (50231428)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
YAMASITA Mitushi Mie University, Faculty of Bioresources, Research Associate, 生物資源学部, 助手 (90158171)
TOYOTA Toshio Japan Condition Diagnosis Technology Laboratory Incorporated, Ltd., President, 代表研究者
HAMAMOTO Toshinori Kyushukyohan, Ltd., Manager, 技術開発グループ, グループ長(研究職)
二保 知也 九州工業大学, 情報工学部, 助手 (60295011)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Project Status |
Completed (Fiscal Year 2004)
|
| Budget Amount *help |
¥8,400,000 (Direct Cost: ¥8,400,000)
Fiscal Year 2004: ¥2,000,000 (Direct Cost: ¥2,000,000)
Fiscal Year 2003: ¥2,900,000 (Direct Cost: ¥2,900,000)
Fiscal Year 2002: ¥3,500,000 (Direct Cost: ¥3,500,000)
|
|---|
| Keywords | condition diagnosis of plant machinery / signal processing / genetic programming / genetic program / symptom parameter / condition surveillance / possibility theory / neural network / ファジィ / 人工知能 / 回転機械設備 / 診断システム工学 / 遺伝的アルゴリズム / 設備診断器 / 回転機械 / 可能性理論 |
|---|
| Research Abstract |
In this study, the contracture method to develop new type of intelligent device and system of condition diagnosis for plant machinery was established based on the results of previous research on condition diagnosis for plant machinery using computer information theory (genetic algorithms, fuzzy theory, neural network, etc.), signal processing theory (wavelet analysis, FFT, etc.) and digital signal processing technology. The result of this study is summarized as follows :
(1)The extraction method of week fault signal from measured signal for condition diagnosis of plant machinery has been established using genetic program and statistical theory.
(2)For distinguishing states of plant machinery by the extracted fault signal, the basic symptom parameters used to express the feature of the signal has been defined, and their sensitivity for distinguishing states has been evaluated.
(3)The method of self-reorganization of the basic symptom parameters for automatically generating optimum symptom parameter for condition diagnosis has been proposed and established using genetic programming.
(4)The automatic learning method using the "intelligent condition catcher", which has been developed by a company with the head investigator, has been proposed and established for the intelligent device and system.
(5)The expression method of degrees of each state for precise and efficient diagnosis using the symptom parameters has been proposed and established by possibility theory and statistical theory.
(6)The prototype of the intelligent device and system has been contracted for trial to carry out condition surveillance and precise condition diagnosis.
(7)The efficiency of the method and the prototype proposed and established in this study has been verified using real field data of plant machinery.
|
