Fuzzy c-means clustering recognition in ultrasonic fault detection in bonding steel plate and rubber material

摘要

The article elaborates the use of ultrasonic in testing the quality of the bonding of a steel plate and rubber. We also analyse information such as the echo signal's attenuation coefficient, harmonic amplitude and frequency, and extract the signal energy, signal duration and the product of singular peak value and quantity, which are regarded as the characteristic values. We use the fuzzy c-means clustering algorithm to make a quantitative analysis on steel plate and rubber bonding quality. The results show that the recognition correct rate of the algorithm is 97.5%, and this method can effectively classify and identify the quality in bonding steel plate and rubber material.