چكيده به لاتين
In general, the identification of structural damage is defined as the identification of local damage and the identification of global damage. For small and regular structures, local damage detection is very good, but for large and complex structures with limited access for the test, it is difficult to identify the damage with the local identification. Hence, in order to identify the damage in the whole structure, especially the specific and complex structures, a set of vibration-based methods for damage identification is presented. The fundamental idea for vibration based damage identification is that the damage-induced changes in the physical properties (mass, damping, and stiffness) will cause detectable changes in the modal properties (natural frequencies, modal damping, and mode shapes). Also, vibration-based methods can be expressed as inverse model updating methods, and solved with optimization algorithms. In other words, the difference between the recorded data from the monitored structure and the analytical model of the structure need to be minimized. Therefore, at first an objective function consisting of the main structure data and the model data is defined, and then it is minimized using an optimization algorithm for structural damage detection. In this research, two new vibration-based methods are presented to identify the location and severity of damage in the engineering structures. The first method involves modifying the central search operator of the ant lion optimization algorithm with the Levy flight random walk in order to minimize the objective function and solve the damage identification problem. The second method is a direct method based on the total modal energy (modal strain energy and modal kinetic energy), to identify the location and severity of damage in the complex structures. In order to evaluate the proposed methods, damage identification is performed on the three numerical examples including a continuous beam, a two-dimensional truss, and a three-dimensional truss. Also, challenges such as the effect of the number of the vibration modes and the presence of noise in the input data are also investigated on the proposed methods. The results show the high ability of the modified algorithm with levy flight random walk and the direct method based on the total modal energy to determine the location and severity of damage, even when using noise polluted data.
