Research On Fault Diagnosis Method Of Wind Turbine Drive System Based On VMD

At present, the wind power capacity of our country is larger than other countries in the world. However, with the increase of installed capacity of wind turbines, wind turbine equipment also appears a series of problems, including wind turbine drive system of high failure rate. Wind turbine drive system failure or accident can cause huge losses which seriously affect the economic benefits of the wind turbine power.This paper mainly introduces the development status of wind power generation and domestic and international research on the fault diagnosis technology, on this basis, analyzes the basic structure and main characteristics of the wind turbine drive system. The Focus is on Fault feature extraction methods used in the wind turbine drive system Fault diagnosis.Variational Mode Decomposition(VMD) algorithm can be divided into constructing and solving variation problems, mainly related to three important concepts: classic Wiener filter, Hilbert transform and frequency mixing. It can simultaneously get each mode estimated carries different center frequencies,which is a group of adaptive Wiener filter bank essentially.It uses non-recursive mode decomposition different from Empirical Mode Decomposition(EMD), avoiding accumulating estimation errors of the envelope line caused by recursive mode decomposition and overcoming the end effect. Therefore, failure analysis method based on VMD has research value.For vibration signal of the wind turbine drive system, this paper uses EMD, Wavelet Packet Transform and VMD to extract fault feature, introduces the basic principles and algorithms decomposition method in detail and applies experimental data to analyze. After extracting fault characteristic of the bearing inner ring, analyzing spectrum of the components, comparing the decomposition effect of three methods, we find that the amplitude of features frequency extracted by VMD is highest which can also overcome modal aliasing effects, so the VMD algorithm is suitable for bearing fault diagnosis.Finally, use VMD algorithm to do wind turbine drive system fault diagnosis according to the actual vibration data, mainly for three representative failures: Gear failure, bearing failure and unbalanced fault. And compare the result of fault diagnosis with EMD, WPA algorithm and find that from the aspects of decomposition effect and the amplitude of fault features, VMD algorithm has its own advantages, and also can avoid aliasing mode. In a word, VMD algorithm has a great research value in the wind turbine drive system fault diagnosis.