Application study of EMD-AR and SVM in the fault diagnosis

Because the non-linear early fault signal of equipment is hard to sentenced to a fault category and the fault degree by traditional fault diagnosis method. In order to solve this problem, they are decomposed into a number of intrinsic mode functions (IMF) with EMD method. Figure out each IMF´s energy entropy and establish the AR model for each IMF´s energy entropy. Finally, the auto-regressive parameters and the variance of remnant were regarded as the fault characteristic vectors and served as input parameters of SVM classifier to classify working condition of the equipment. The rolling bearing analysis experimental results show that this approach is good.