The Measurement And Analysis System Of Electric Machine's Vibration Based On Virtual Instrument

The electric motor is one sort of the most important electric equipments in modern manufacturing. Its failure has serious impact on the manufacturing. The environmental protection requires controlling the electric motor's noise. Measuring and analyzing the vibration of electric motor provides a way to detect motor's fault and reduce noise. It is vital of importance to develop a measurement and analysis system.Virtual instrument technology is widely used in measurement due to its development in the last 20 years. Compared to traditional instrument, Virtual instrument has many advantages, such as better performance. LabVIEW is one of the best virtual instrument software platforms.This thesis implemented a measurement and analysis system for electric motor's vibration. The main content of the thesis includes the followings:1. A virtual instrument based measurement and analysis system is developed. The function of this system covers the vibration signal's acquisition, display, and process and, data management.2. The study of signal processing methods, such as FFT, time-frequency transform, wavelet analysis, and its application in vibration signal processing are conducted. ICA (Independent Component Analysis) and its application in vibration source separation are discussed. The optimized parameter of various signal processing methods is put up.3. The methods of feature extraction and selection related to electric motor's vibration are extensively studied in this thesis. A LS-SVM based fault diagnosis system is designed to demonstrate the effectiveness of SVM.4. A new classifier is put up to solve the problem that it is difficult to get enough of samples about the motor false status except the normal condition. This classifier is a promising method.