Improved Genetic Algorithm for Structural Damage Detection

Structural damage detection consists of determining the location and severity of damage in civil structures by using measured parameters. Genetic algorithm is a meta-heuristic computing method for finding approximated global minimums in large optimization problems and has the advantage of probabilistic hill climbing. In this work the improved genetic algorithm is employed to solve damage detection problem in truss type structures using vibration data (natural frequencies and mode shapes). The formulation of the objective function for the optimization procedure is based on natural frequency and vibration mode shapes. Present structural damage-identification scheme is confirmed and assessed using a Finite Element Model formulation (FEM) of truss structure. Results are presented in tables.