Rolling bearing fault diagnosis method based on wavelet packet and GWO-SVM

The invention relates to a rolling bearing fault diagnosis method based on a wavelet packet and GWO-SVM, and belongs to the field of mechanical fault diagnosis. The method comprises the following steps: measuring different vibration signals of the rolling bearing in an inner ring fault, an outer ring fault, a rolling body fault and a normal state by using an acceleration sensor; performing three-layer decomposition on the signals through a wavelet packet; calculating the energy of eight nodes on the third layer as characteristic data of the signals; after extracting the characteristic, inputting the characteristic parameters into a Support Vector Machine (SVM) optimized by the gray wolf algorithm for training, and obtaining a fault diagnosis model based on the GWO-SVM; and finally, inputting test data into the trained model to finally obtain the fault type of the test data. Experimental results show that the accuracy of fault recognition by the gray wolf algorithm is significantly improved. By taking fault di