Fault diagnosis based on Walsh transform and support vector machine

Recognition of shaft orbit plays an important role in the fault diagnosis. In this paper, a novel recognition method for the shaft orbit based on Walsh transform and support vector machine is proposed. In the method, distance vector between the point on the shaft orbit and its center is first calculated. Then, the distance vector is transformed by Walsh matrix, and the Walsh spectrum obtained has property of invariance to rotation, scaling and translation. In the end, the Walsh spectrum, viewed as the feature of shaft orbit, is trained and tested by means of support vector machine. In addition, a comparison with the previous methods is performed, and experimental results are encouraging, which fully demonstrates the effectiveness and superiority of the proposed approach.