Modal Strain Energy Based Damage Detection Using Multi-Objective Optimization

Modal strain energy has been reported by researchers as one of the sensitive physical measures that can be used as a damage index in structural health monitoring. Inverse problem-solving based approaches using single-objective optimization algorithms are also one of the promising damage identification methods. However, integration of these potential methods is currently limited with partial success in the detection of structural damages due to errors and noises. In this study, a novel damage detection approach using hybrid multi-objective optimization algorithms is proposed to detect multiple damages in a 3-dimensional steel structure. This study developed an approach to overcome the shortcomings of the single-objective genetic algorithm based approaches using multi-objective formulations for minimizing errors of damage indices between actual damaged structures and simulated damages. The performance of the proposed hybrid multi-objective genetic algorithm is compared to that of traditional single-objective optimizations based approaches. This study accurately detects the location and extent of induced multiple minor damages of the laboratory 3-dimensional steel structure.