Structural Damage Detection Based On ACO Algorithm

In the field of structural damage identification and structural health monitoring, some theories have been developed and broad application prospect are expected. As research hot spots and difficult problems in the field, structural damage identification methods based on vibration characteristics have attracted researchers' wide attention and become a very active branch in mechanics and civil engineering. However, there are still some difficulties when they are applied to the real structures.In this thesis, the Ant Colony Optimization (ACO) algorithm is introduced to the structural damage detection field in order to provide a new exploring spot for structural damage detection method. First of all, the common methods applied to the structural damage detection field are summarized. Some issues on structural damage detection are converted into a constrained optimization problem, which is then hopefully solved by the ACO algorithm. Secondly, the problem of structural damage detection are modeled in mathematics, a definition on optimization problem is provided, the computation principle and procedures for four optimization algorithms, such as swarm intelligence, are introduced as well. Thirdly, after having presented the ACO algorithm, it focus on the continuous ACO algorithms, seven test functions are adopted to test its performance. Fourthly, the continuous ACO algorithm is applied to the structural model updating and structural damage detection. Based on the numerical simulations for single and multiple damages of a 2-story rigid frame and on the experimental study on damage detection of a 3-story building model, some illustrated results show that the ACO algorithm is very effective for structural damage detection. The ACO algorithm can not only locate the structural damages but also identify the severity of damages. Regardless of small damage or multiple damages, the identification accuracy is very high and noise immunity is better, which shows that the ACO-based algorithm is feasible and effective for structural damage detection. Finally, a few conclusions are made, some possible research directions are suggested, in particular, the potential issues on converting the structural damage detection into constrained optimization problems are discussed.