Structure Damages Detection Based On Genetic Algorithm with Experimental Validation

This paper describes a procedure about detecting of beam-like structural damage using a genetic algorithm with experimental validation. Firstly, the decreases of element stiffness is employed as the target function, then, a set of finite element model (FEM) of simply supported beam are founded for getting the dynamic parameter (element stiffness), next, a genetic algorithm is trained using these parameters as the target function. Finally, a group of steel beams are used to test the genetic algorithm for the detection of cracks damage.The effects of the damage detection based on the agenetic algorithm are also discussed though the FEM and the experimental.