Degradation feature extraction method for piezoelectric ceramic of ultrasonic motor based on DCT-SV cross entropy

Crack on piezoelectric ceramic is the main reason leading to failure of ultrasonic motors. A novel degradation feature extraction method based on discrete cosine transform (DCT) -singular value (SV) cross entropy was proposed in this paper. In order to improve the correlation with the crack, the DCT coefficients with the property of energy aggregation, were used to extract fault information. To avoid the influence of human factors in traditional DCT de-noising method, a matrix composed of DCT coefficients was constructed, and the SV cross entropy of the matrix was taken as the degradation feature for ultrasonic motor. A numerical simulated noise was added to the measured signal to verify the anti-noise performance of the feature. Analysis of the experimental results demonstrates that the proposed DCT-SV cross entropy is feasible and effective in indicating the degradation of piezoelectric ceramic in ultrasonic motor.