Development of an experimental software for rub-impact fault recognition research

Abstract

Rotor and stator rubbing is a common fault of rotating machinery, for instance, aircraft engine, currently, numerous methods are presented to address the issues of Rub-impact fault. For the purpose of effective research for the different diagnosis methods, a experimental system for Rub-Impact fault diagnosis research is achieved in this paper, with the mixed programming of Matlab and VC++ applied. As an illustration, three methods, such as Linear-SVM, RBF-SVM and BP network, are embedded and the respective diagnosis results are analyzed. In addition, a brief introduction of mixed programming with Visual C ++ and MatLAB and the system architecture are given in this paper. The application of software platform gives the results that the system presented is practical and effective in fact.