DAMAGE IDENTIFICATION OF PRESTRESSED CONCRETE BEAMS BASED ON MODAL STIFFNESS

This paper expounded that modal stiffness are of great significance for damage identification． It was proposed BP and PNN neural network for damage identification based on this modal parameter． The result，via experiment analysis on 11 prestressed concrete beams with multistage damage status，shows that the theoretical method based on modal stiffness is hardly practical in engineering application due to many factors such as noise and boundary condition etc，while the neural network can be effectively applied to identify damages with high precision． Moreover，which can offer a new idea for damage identification researches．