Research On Feature Extracting Technolegy And Feature Extracting Software System For Faults Of Rotating Machinery

Extracting features of the rotating machine faults is one of the most important processes in the rotating machine faults diagnosis. Based on the "Resarch on New Method for Rotating Machine Faults Diagnosis Based on Independent Component Analysis "(National Nature Science Fund Project, No: 50205025), the "Research on Technique for Fault Diagnosis with Mechanical Noise Based on Blind Source Separation"(Nature Science Fund Project of Zhejiang Province, No: 5001004), and the projects of Fujian Petrochemical Co. and Quzhou Juhua Co. Ltd, we studied the technology of extracting features from vibrational signals of rotaries. We studied the traditional technology of extracting features of rotating machine faults, such as the numerical data magnitude analyzing, time area analyzing and several classical feature extracting technology based on FT; discussed the several signal processing methods of Short Time Fourier Transformation (STFT), Wavelet Transformation (WT) and the empirical mode decomposition (EMD) based time-frequency analysis and their application in the rotating machine faults feature extracting process. We made some experiments to compare these methods and testify our research results. In the end, we used Matlab and C++ mixture programming languages to develop a rotating machine's faults feature extracting software system based on FFT, STFT, Wavelet Transformation and EMD. The details are studied as follows:Chapter one briefly introduced the general situation of vibration monitoring and faults diagnosis of the rotating machine. The developing and the current situations of the technology of extracting feature of rotating machine's faults were summarized. At last, the objective, main contents, general structure scheme of this dissertation were presented.Chapter two studied the traditional technology of extracting features from rotating machine's faults, such as the numerical data magnitude analyzing, time area analyzing and several classical feature extracting technology based on FT, and then summarized the characteristic of these methods.Chapter three researched the basic ideas of Wavelet Transformation theories, and discussed the specialties in using the virtue of Wavelet Transformation to extract features of the rotating machine's faults. Verified the theories and methods of WT through analyzing the simulate signals and the real vibration signals of the rotating machine with WT.Chapter four researched the basic ideas of EMD theories and arithmetic, then discussed the important issue of EMD's end effect. Studied the direct short data extending technology to weaken the influence of end effect. At last, analyzed a real vibration signal of the rotating machine through the method of EMD and the conclusion was deduced that the EMD-based time-frequency analysis method is valid in the field of signals processing in rotating machine.Chapter five compared the EMD-based time-frequency analysis method to FFT, STFT, Wavelet Transformation, used these methods to signal process and extract the features of four faults of Bently rotor machine, and concluded some research results.Chapter six used Matlab and C++ mix programming languages to built the rotating machine's faults feature extracting software system based on FFT, STFT, Wavelet Transformation and EMD. At last the basic components and functions of the software are illustrated.Chapter seven is the final part of this paper, summed up the work of the dissertation andpresents conclusion and prospect.