Performance Investigation of Metaheuristic Niched-Pareto Genetic Algorithm for Imperfection Assessment of Structures

Abstract

Structures depending on the type of their application are subject to different types of loading, such as wind, earthquake and explosion loads. On the other hand, in order to increase the useful lifetime of the structures, optimal operating conditions must be provided. Inappropriate utilization, excessive loading, changes in structural stiffness due to the material imperfection will cause the structure to come out of its design conditions and leads to unsafe structure. Therefore, on time assessment of imperfection for retrofitting and rehabilitation in order to increase the useful lifetime of the structure is very important. This article presents a novel and robust approach for assessment of imperfections in the structure using the concepts of vibration analysis and finite element method. A multi-objective Niched-Pareto Genetic Algorithm was used for computation of geometric position and intensity of imperfection in a steel continuous beam structure as well as a cantilever beam. The results of the imperfection assessment in continuous steel beam demonstrate that the proposed method has great potential in localization and quantification of imperfection in the multi-stage scenario. In addition, the results of this study for cantilever beam were compared with the same experimental results reported in the literature. These results demonstrated that the computational error in the proposed method was less than 0.06, while the error of similar experimental results has been reported as 0.3. Finally, the results of this research, demonstrate that the metaheuristic Niched-Pareto Genetic Algorithm performs very well and therefore this algorithm is recommended as a robust and practical approach for imperfection assessment of structures.