DAMAGE IDENTIFICATION OF COMPOSITE STRUCTURE USING HILBERT-HUANG TRANSFORM AND ARTIFICIAL NEURAL NETWORK

摘要

An active damage detecting approach based on Hilbert-Huang Transform spectral analysis and artificial neural network technology is developed for detecting the local delamination in a composite laminates. A piezoelectric ceramics wafers array, bonded on the composite laminate plate surface, is used as the actuators and transducers to generate and receive the ultrasonic Lamb wave and to detect the damage in the plate. A new time-freq ency domain signal process method, Hilbert-Huang transform, is employed to analyze and extract the characteristics of the damage from the signal of sensors. The damage characteristic vector is defined and extracted according to the instantaneous energy spectrum and the margin spectrum of the signal. With the damage characteristic vectors as the inputs, the BP neural network is used to identify the position and the area of the damage in the structure.An active experimental diagnosis system is established. And the results of the research of the experiments on composite laminate show that the approach proposed in the paper could effectively detect the simulating delamination of the plate.