抄録
The present paper is about the improvement of the delamination identification of the CFRP structure by the electric potential method via the discriminant analyses. In general, the relation between the damage and measurement values is identified by solving an inverse problem. The inverse problem in damage identification is one of the optimization problems involving the minimization of estimation error, and is usually solved by common optimization tools such as neural network and genetic algorithm. However, trial and error processes are generally required to derive the optimum solution using these methods, and these processes demand much calculation and human cost. In the present paper, delamination identification of CFRP beam via discriminant analysis using Support Vector Machine is conducted. The present study adopts an electric potential method for monitoring the structural integrity of laminated composites. As a result, it was shox that this method is effective for identification of damages.
