Fault diagnosis based on rough set and Support Vector Machines for aero hydraulic pump

Taking into account that fault samples of the aero hydraulic pump were insufficient and the fault data thereof were characterized by imperfection and redundancy, the reasons and characters of gap increase of the piston ball-head of a hydraulic pump were analyzed, the frequency domain signals of fault pumps and those of normal pumps was compared, and a fault diagnosis method based on rough set and Support Vector Machine (SVM) was put forward. The rough set was utilized to reduce the fault characteristic value and eliminate redundancy, thereby finding the minimal attribute to describe system fault characters on the premise of unchanged classification quality. And then, the sample data disposed by rough set were used to train SVM to realize fault diagnosis of small samples. The experiment result shows that the method, adopting both rough set and SVM, is suitable for high-precision fault diagnosis of the aero hydraulic pump.