Research Of Fault Feature Information Extraction Based On The Empirical Mode Decomposition

Fault diagnosis is a science of identifying the running state of device by studying the varying information of the device running state. Signal processing is the core of fault diagnosis technology. In nature, almost every signal is non-stationary. Some popular methods have been analysis, such as time-frequency analysis, Wigner-Vill distribution and wavelet analysis, but they have some faults. An American-Chinese N. E.Huang firstly put a new method forward—Empirical Mode Decomposition (EMD) method.There is an involved end issue in the course of getting two envelops of the data using spline interpolation. The cubic spline will swing widely if the end issue is left unattended, and it will affect the veracity of the instantaneous characters' extracting. There are some methods to solve the problem, they have confined the large swings successfully .In the paper, it has brought combining 'symmetrical continuation' with 'adding extremes', and has provided its correctness by making use of datum.The method is valid in the field of fault diagnosis character extraction, by carrying out the improved empirical mode decomposition method. Then, it illustrates the effect of the EMD-based filtering technique. By the simulation and application, it shows that EMD-based filtering technique is very useful in extracting fault diagnosis character.We use fault character information extraction methods to signal process and extract the features of two faults of Bently rotor machine, and conclude some research results. With MATLAB and VB mixture programming languages, the fault diagnosis system is visual.