Application of improved FCM clustering algorithm in felting quality ultrasonic detection of composite material

Taking ultrasonic detection signal of composite plate material felting quality for the research subject, the characteristic of time and frequency domain of echo signal is comprehensively analysed, and attenuation coefficient,signal energy and the product of singular wave peak value and the quantity are extracted as characteristic values; with the help of IFCM(Improved Fuzzy C-Means) clustering algorithm, the felting quality of the composite material has been sorted in fuzzy clustering. Finally, fuzzy clustering classification is converted to the accurate classification according to the maximum membership rule. The experiment shows that the adoption of this method results in a high accuracy and a short convergence time.