عنوان مقاله [English]
Engineering structures are prone to damage over their service life as a result of natural disaster so that damage spreading may lead to many casualties. In order to prevent these catastrophic events, early damage detection must be carried out. By considering these issues, numerous structural damage detection methods have been proposed by many researchers in the last few decades. Among all sorts of methods developed for damage detection in structures, vibration-based methods due to their simplicity and applicability are highly favored by many researchers. The basic conceptual of the vibration-based methods is that modal parameters (natural frequencies and their associated mode shapes) are functions of the physical properties of the structure (mass, damping, and stiffness). Therefore, changes in the physical properties will cause changes in the modal properties. A class of vibration-based methods is identified and damages are quantified using the model updating approach. In these methods, an objective function defined in terms of the discrepancies between the analytical model and real structural system is minimized as an optimization problem. In this paper, a novel model updating method is presented based on a structure’s main modal parameters (natural frequencies and their corresponding modal shapes). For this purpose, a hybrid vibration-based objective function is proposed to minimize the differences between the structure’s properties and the analytical model. A penalty function is integrated into the objective function to reduce the effects of noise in damage detection and uncertainties in the assessment procedure. The Teaching-Learning-Based Optimization (TLBO) algorithm is applied to solve this problem as an optimization problem. This algorithm is inspired by the traditional learning process of students in school. The two main stages of this algorithm are the effect of the teacher’s knowledge on student learning by the convergence strategy and students learning from each other by the divergence strategy. To evaluate the applicability of the proposed objective function in detecting the location and intensity of the damage, three numerical cases are considered. These cases include an 8-story shear frame, a continuous beam, and a spatial truss. Different challenges such as the effect of noise on measured data and the effect of the penalty-function on results of damage detection were considered. Furthermore, a comparative study is investigated between the proposed objective function and three other objective functions developed based the main model parameters. The results demonstrated that the proposed method is a reliable and stable technique in damage prognosis in structures.