Study On Bearing Fault Diagnosis Of Asynchronous Motors Based On Wavelet Packet And EMD

Motor is a kind of driving machinery which is widely used in modern society. With the progress of modernized science and technology and constant development of production system, the motor is playing an increasingly important role. Induction Motor is mostly used in the production and lives. It has many advantages, such as simple structure,low cost,high reliability,durability and convenient maintenance. It can not only damage motors itself but also affect the normal work of the whole production system and even endangers people if motors doesn't work and are in faults. Furthermore, this will result in huge economic loss and bad social impact. Thus it is urgent for diagnosis of motor fault.Almost half of the motor faults are related to the bearing fault, because the bearing is the one of electrical components which works in the worst condition. It is used to withstand load and transfer load. As a result of loading, installation, lubrication condition and other factors, running after a period of time, it will produce a variety of different types of faults. Therefore, motor bearing is a relatively weak link, and its operational status often directly affects the performance of the entire machine.In this paper, the author took motor bearing as the research object, carried out an analysis of inner fault, ball fault and combined failure of the two, and elaborated the fault mechanism of motor. The author also analyzed the reason which results in breakdown and concluded frequency performance of several kinds of common faults. During the trial, we collected vibration signals of the motor fault to analyze.In the fault detection technology, signal fault characteristics of analysis and extraction are the key to the fault diagnosis and direct relation to the accuracy of fault diagnosis. Motor fault signal is unsteady. Fourier transform cannot effectively extract the characteristics of the fault motor, Wavelet transform has good time-frequency localization and the signal can be divided in any band, this paper took the wavelet packet transform to extract fault characteristics information of signal. What's more, wavelet packet has a good inhibitory effect to the noise signal and de-noising properties are very obvious, so it was proved that wavelet packet has the most superior effect to the signal de-noising by comparing the de-noising effect of the wavelet and wavelet packet.Empirical mode decomposition is a new signal processing method which based on the signal of local characteristics. Using it, the author can obtain a time-series——Intrinsic Mode Function, which makes instantaneous frequency meaningful. It is particularly suitable for the analysis and processing of non-linear, non-stationary signal, and we can acquire the signal characteristics of expression and information.In this paper, the author combines the advantages of the both, proposes a new method for bearing fault diagnosis with the combination of wavelet and EMD. This method can highlight the characteristics of the data generated by vibration signal with the motor bearing fault condition and extract them. It overcomes the limitations of the fast Fourier transform. Furthermore, the author deals with the signals of inner fault, ball fault and combined failure of the two which are made during the trial, extracts the characteristic of fault and categorizes the fault. Through the method we can solve the problem of fault diagnosis of motor bearing better.