A fault diagnosis method of wind turbine gearbox based on CNN-ISABO-SVM

Aiming at the problem that the early fault signal of wind turbine gearbox is weak and difficult to extract effective features, this paper proposes a fault diagnosis method based on convolutional neural network (CNN) and improved support vector machine (SVM). The S-transform is used to convert the one-dimensional gearbox vibration signal into a two-dimensional feature map containing time-frequency characteristics, and the CNN is constructed to extract features from the time-frequency map. The Latin hypercube and golden sine strategies are introduced to improve the Subtractive average Optimization Algorithm (SABO), and the improved SABO algorithm is used for parameter searching of the SVM, and an ISABO-SVM classifier is designed to classify the extracted fault features and derive the fault diagnosis results. The experimental results show that the CNN-ISABO-SVM wind turbine gearbox fault diagnosis model proposed in this paper has higher accuracy compared to the CNN-SVM and CNN-SABO-SVM fault diagnosis models.