چكيده به لاتين
Abstract:
In recent years, investigation on vibration-based damage identification has gained more attention. The main idea of these methods are based on this fact that the changes in physical characteristics (mass, damping and stiffness) leads to detectable changes in modal properties (natural frequencies, modal damping and mode shapes). Due to some defects in convergence speed of previous method in the literature, there is an increasing demand for methods with high performance in damage detection problems. In chapter 1, theory, assumptions and necessities of the current research were introduced. Afterwards, a literature review of damage detection methods has been provided in chapter 2.
In chapter 3, the concepts about grey wolf optimization (GWO) and modified hybrid particle-swallow swarm optimization (MHPSSO) have been investigated. Afterwards, the first method has been proposed which its cost function is based on direct data-fitting strategy using modal residual force (MRF). In the second proposed method, modal residual force and modal assurance criterion (MAC) have been used in order to define a novel cost function. In both of these methods, the grey wolf optimization has been employed in order to find the optimum solution. Using MRF and optimization at the same time has helped the process related to finding the location and severity of damage simultaneously. In the third method, the generalized flexibility matrix and kernel principal component analysis have been utilized in order to develop a method based on the cost function of the second method (based on modal assurance criterion), cross-section stiffness (stiffness defined with a damage on the element) and the modified hybrid optimization algorithm.
In the chapter four’s examples, the efficiency of the proposed methods has been investigated. The first three examples are related to the first two methods. These examples are benchmark structure, 2-dimensional truss and planar four-story shear frame, respectively. In the next section which is related to the third proposed method’s examples, a 2-dimensional finite element frame with three element is used. Subsequently, two examples with more elements and complicated damages have been utilized which are 11-story frame and 3-dimensional truss, respecively. The noise effect on natural frequencies also has been considered. The results show the viability and efficiency of the proposed methods (considering noise effect).
Keywords:
Damage detection, grey wolf optimization, modal residual force, modal assurance criterion, generalized flexibility matrix
