Fault Diagnosis Of Gear-box Based On Hilbert-huang Transform And Support Vector Machine

As an important part of mechanical equipments, Launched a research on monitoring and fault diagnosis of gear-box has strong practical significance. In this paper, Simulation experiments of gear box common fault were done to extract the feature of signal with HHT.SVM classifier was used on the gear box of the type of job status and fault classification, and gets a better effect.The process of machinery fault diagnosis includes the acquisition of information and extracting feature and recognizing conditions of which feature extraction and condition identification are the priority. When the fault of gear box occurs, the vibration signals often show non-stationary characteristics. The EMD method which is proposed in this paper is based on the local characteristic time scale of signal and decomposes the complicated signal into a number of Intrinsic Mode Functions (IMF).Each IMF components including the different features of time scale composition. In addition, the time scale is changing from small to large. Thus, each IMF component contains different frequencies which change from high to low. The basic theory of EMD is applied to fault diagnosis of gearbox in this paper. Do the HHT transform to some IMF components which contain abundant failure information to obtain HHT Marginal spectrum, then to initially achieve fault diagnosis of gear-box.Support vector machine (SVM) as a new learning machine, it is one of the statistical learning theories, which developed on the basis on the statistical learning theory. Currently, a support vector machine has become a powerful tool in solving nonlinear classification problems. Besides, SVM has Small dependence on experience, and has better generalization and guarantee the local optimal solution is exactly the global optimal solution. SVM now has been widely used in pattern recognition. The experimental results demonstrate the proposed diagnosis approach in which EMD Energy feature extraction and SVM are combined is effective.