---

概要

Abstract: When multi-faults state occurred in a plant rotating machine, because the relationship between fault types and vibration signals measured for the diagnosis are much more complex than the case of single fault, moreover, the condition diagnosis information from vibration signals is ambiguous, especially at early stage, the diagnosis for the multi-faults is more difficult. In order to resolve the problem, this paper proposes a new precise diagnosis method for rotating machinery using support vector machine (SVM) and fuzzy theory. Firstly, the optimal classification-line for distinguishing two states can be obtained using two SPs and SVM. Then the synthetic symptom parameter (SSP), that has higher sensitivity for the diagnosis, can be obtained according to the optimal line of classification. Finally a fuzzy diagnosis method is proposed to discriminate states of rotating machinery using the ambiguous diagnosis information. The practical diagnosis examples of bearing with multi-faults are provided to verify the effectiveness of the proposed method.

---

全文掲載のPDFファイルダウンロード

• PDFファイルダウンロード