Fault Diagnosis Of Motor Rotor Unbalance Based On Improved Hilbert-Huang Transform

As the main executive component in the modern industrial production, whether the electromotor works normally or not will have a direct influence on the quality of the whole production process. Rotor unbalance fault is a relatively common type of motor faults. It can cause serious damage to motors, but it cann’t be detected easily. Therefore, for the purpose of helping dectect the fault timely in its early stage, and preventing the futher deterioration of the fault, the research on the fault diagnosis technique of rotor unbalance, has important theoretical significance and practical economic value.Firstly, the current situation of the research on the diagnosis technology of rotor unbalance fault and the research status of vibration analysis methods were described, which lay a theoretical foundation for using Hilbert-Huang Transform to detect and diagnose the rotor unbalance fault. Hilbert-Huang transform includes two processes which are the empirical mode decomposition(EMD) and the Hilbert spectral analysis. EMD is an adaptive method, which can decompose nonstationary and nonlinear signals very well. The funtion of the EMD is to obtain the desired intrinsic mode function(IMF). Then the signal spectrum and marginal spectrum can be obtained by using Hilbert spectral analysis to analyze the IMF. By doing this, people can analyze the fault signals of rotor unbalance.Secondly, in order to solve the endpoint effect in the empirical mode decomposition, an improved method that combines the mirror extension with the support vector regression is proposed based on analyzing the traditional methods. In the improved method, the support vector regression method is applied to predict extreme points on both ends of the extreme points of the original signal, and then the mirror extension method is applied to determine the position of the predicted extreme points. The improved method can solve the inaccurate prediction on the long data sequence by using the support vector regression method separately, and the problem which the boundary of the short time sequence is not the extreme point by using the mirror extension method separately. By the simulation,it can make qualitative analysis between the improved method and the traditional endpoints extension methods. The results show that the improved method can suppress the endpoint effect very well. Then, using the improved Hilbert-Huang transform to diagnose and analyze the electromotor rotor unbalance fault, the characteristics of fault signals can be accurately extracted from intrinsic mode function and marginal spectrum.Finally, combining with the improved Hilbert-Huang transform method, a fault diagnosis system of rotor unbalance based on Lab VIEW was designed. By the analyses of experiments, this system can effectively diagnose rotor unbalance faults.