A modified Artificial Bee Colony algorithm for structural damage identification under varying temperature based on a novel objective function

• Published in: Applied Mathematical Modelling Vol. 88; pp. 122 - 141
• Main Authors: Ding, Zhenghao, Fu, Kangsheng, Deng, Wu, Li, Jun, Zhongrong, Lu
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.12.2020; Elsevier BV
• Subjects: Artificial Bee Colony algorithm; Damage detection; Environmental effects; Evolutionary algorithms; Measurement noise; Modal data; Modulus of elasticity; Optimization algorithms; Resonant frequencies; Robustness (mathematics); Search algorithms; Structural damage; Structural damage identification; Swarm intelligence; Temperature effects; Temperature variations; Structural damage identification; Artificial Bee Colony algorithm; Measurement noise; Temperature variations; Modal data
• ISSN: 0307-904X; 1088-8691; 0307-904X
• DOI: 10.1016/j.apm.2020.06.039

•A modified Artificial Bee Conoly algorithm is propsed for structural damage identification.•The effect of environment is taken into account during structural damage identification.•The proposed method can be applied to space truss structure and 2-D plate structure.•The proposed method obtains accurate identification results with noisy measurement data. This paper presents a modified Artificial Bee Colony algorithm for structural damage identification. Meanwhile, the effect of temperature variation is considered and the change of temperature will lead to the alteration of Young's modulus of material. A novel objective function is proposed as the combinations of the partial mode shape curvature data, alterations of natural frequencies, and a sparse penalty term. Such an objective is found to be sensitive to structural damage while not sensitive to environmental effects. On the other hand, To render the standard Artificial Bee Colony algorithm more powerful and robustness, two local search strategies are introduced into the employed and onlooker bee phase of the Artificial Bee Colony algorithm, respectively. Two numerical examples and a laboratory verification are employed to verify the efficiency and advantage of the proposed algorithm. The final results show that the present algorithm could yield more satisfactory identification results compared with other state-of-the-art evolutionary algorithms, even high-level noise and temperature variation are considered; and the proposed novel objective function is more sensitive to structural damages, compared with the traditional mode-shape-based objective function.