Fault diagnosis of pitting corrosion for gears under variable loads based on multi-dimensional features and grey wolf algorithm with support vector machine

In order to deal with the difficulties of extracting gear fault information and identifying fault types, a method based on the multi-dimensional features and the Grey Wolf Algorithm with the Support Vector Machine (GWO-SVM) is developed for pitting fault diagnosis of gears under variable loads. First, a multidimensional fault feature extraction method combining the Empirical Mode Decomposition (EMD) algorithm, the EMD energy entropy and the fractal box dimension is proposed; then, the Gray Wolf Optimizer is applied to optimize the parameters of the support vector machine, from which the GWO-SVM fault identification is proposed; Finally, the fault feature extraction method and the fault identification method are integrated. A rotating machinery fault diagnosis test bench is constructed to carry out feature extraction. From the analysis the effectiveness of the method proposed in this paper for diagnosing pitting faults of variable loaded gears is verified. The results show that the accuracy of the proposed method is approximately 97%, which is better than that of individually extracting the time domain and frequency domain features of the fault signals.