In this thesis,use acceleration sensor to get headstock′s vibration signal,use the decompounding of wavelet package to get fault feature vector,use eigenvector′s training to transmit BP Neural Net forward,finally set up the sorter of gear s ruming state to identity gear fault.

This article proposed a new method for Extraction fault feature vector of Analog Circuit——Extraction some nodes\' voltage value and magnitude of current to Construction the fault feature vector of Analog Circuit.

This paper analyzes and abstracts the fault feature of the pressure signal generated by operating gear pump which is part of the loading machine hydraulic system when oil box is lack of oil through digital signal filter,fft.

Research On Extraction of Fault Feature of Hydraulic System in the Loading Machine;

Longitudinal zero-sequence voltage and fault phase current were analyzed as stator winding’s inter-turn short-circuit stable fault characters,mathematical model of which were bui.

Extraction of the Fault Features of Ball Bearings Based on Bispectrum Analysis;

Method of fault feature extraction for analog circuits based on EMD

In order to get a characterized wavelet with expected properties of fault features extraction,by using the lifting scheme(LS),a new wavelet is constructed.

Analysis of the fault characteristics of a rotor with a transverse crack;

The fault will result in fault characteristic component in stator current.

The fault characteristics of transient travelling waves are described using the dyadic discrets wavelet transform (DDWT) to present the travelling waves.

With FEM used to simulate the rotor system on which the looseness fault is occurring,the fault characteristics are obtained as follows.