Improving the Performance of Structural Damage Detection Methods Using Advanced Genetic Algorithms

Abstract

A frequency response function-based damage identification method is presented that accurately identifies both the location and severity of damage in structural systems using a limited amount of measurement information. Damage is identified by minimizing the error between measured and analytically computed frequency response functions obtained through finite element model updating. The impact that the type of genetic algorithm representation has on performance is evaluated for a fixed representation and an implicit redundant representation, which simplifies the search by exploiting the unstructured nature of damage identification. The performance of the proposed damage identification method is evaluated for beam and frame structures that consider different damage scenarios and measurement layouts. The impact of measurement noise on performance is also investigated. The damage identification method developed using the implicit redundant genetic algorithm provides greater accuracy in identifying the location and severity of damage in all case studies even in the presence of noise. For larger frame structures, the implicit redundant genetic algorithm performed well, while no valid results were obtained using the fixed genetic algorithm representation.