Application Of Wavelet Analysis In Fault Diagnosis Of Rotating Machinery

With the rapid development of modern industries, the safe and stable operation of the equipment increasingly high demand, fault diagnosis technology has been widely applied. At the same time, traditional diagnostic techniques have been very hard to meet the needs of industrial production. In this thesis, a new method of time-frequency analysis — wavelet analysis is applied in fault diagnosis of rotating machinery.There is a great deal of non-stationary signals in machinery operation. But traditional time domain analysis and frequency domain analysis have many limitations with non-stationary signals. Wavelet transform bases on base function, learns the characteristics of triangle base of Fourier Transform and window function of Short-Time Fourier Transform, and forms oscillatory and damped base function. It is praised as a "mathematical microscope" because of its multi-scale and time-frequency characteristics and becomes an important tool for dealing with non-stationary signals.First of all, this thesis describes the background, the development of fault diagnosis and its current development all over the world. And then it introduces ordinary fault of rotating machinery and its mechanism briefly, which helps us know what is need for fault diagnosis and how to diagnose. In the third chapter, there is an overall design of the whole system and brief introduction about every module. At the same time, this thesis compares wavlet analysis with some traditional methods of fault diagnosis, and concludes the necessarity of using wavelet analysis for fault diagnosis of rotating machinery. The last two chapters detail the system design and implementation of each of the two core modules. As a result, practice has shown that the method can improve the accuracy of fault diagnosis, and achieved good results. Meanwhile, the results could also provide data for expert system or intelligent diagnosis.The issue stems from the background of the need of condition monitoring of key equipment — tobacco turbine in some petrochemical corporation. To ensure that the machine doesn't go wrong, real-time monitoring and fault diagnosis based on computer are raised, which reduce unnecessary maintenance and parking and increase automation and adaptive capacity. The thesis conducts a comprehensive study of rotating machinery for the adaptability of the developed product. And in application, in order to ensure the applicability and reliability of the system, advanced sensors and measurement modules are used.