Precise Damage Identification Of Structural Based On Particle Swarm Optimization

As one of the important methods in the field of structural health monitoring (SHM), vibration-based structural damage detection (SDD) methods have attracted more and more attention and engineering researchers. However there exist some difficulties and shortcomings when the SDD theories are applied to the practical structures. Particle swarm optimization (PSO) is a kind of swarm intelligence optimization. For its rapid convergence and simple implementation, PSO has been widely used in function optimization, neural network training, fuzzy system control and other fields, and it has been applied to engineering applications in recent years. In this thesis, these two different technologies in two different areas are combined and improved in order to develop new methodologies for the SDD as well as to extend the applications for the PSO.Based on the previous work, we did further investigate PSO-based SDD methodologies here. Firstly the vibration-based SDD techniques are summarized, and PSO algorithm is introduced next. In order to solve the problem that the accuracy for weak and multiple damage identification is not so high using the existed methods, some new objective functions based on modal flexibility index and modified modal assurance criterion are proposed for the SDD. Finally the effectiveness and robustness on new objective functions used for the SDD are evaluated and compared with the conventional one through numerical simulation and experiments in laboratory. In the numerical simulation, less damage information and a spatial truss structure are adopted to identify the damage. The advanced LMS Test. Lab system is adopted to obtain the modal parameters of structures through the experimental modal analysis, in which a steel beam hinged at both ends is used to do the experiments of the health and damaged structures. The illustrated results from both numerical simulation and experimental verification show that the new objective function combined with PSO algorithm is more excellent than the conventional one, especially suitable for detection of weak and multiple damages occurred in structures.