Bridge damage identification by combining modal flexibility and PSO methods

Structural damage detection is a vital step in the structural health monitoring. It is often converted into a constrained optimization problem finding the optimal damage distribution that simultaneously minimizes the objective function. First, two objective functions are defined. One is defined as minimizing the sum of differences between the modal data before and after damage in traditional way. The other is newly defined based on modal flexibility, which is combined with another function able to predict damage location. Secondly, an improved particle swarm optimization (PSO) algorithm is developed based on the macro economic strategies and used to solve the multiple-objective optimization problem on bridge damage identification. Thirdly, a simply supported bridge and a continuous bridge beam, as the models of numerical simulation, are adopted to evaluate the efficiency of the proposed procedures mentioned above. The illustrated results show that the weak and multiple damages on the bridge models can be well identified by the proposed approach, which show that the procedure is very promising for locating and quantifying damaged elements of bridge structures and considerably improves predictions based on the modal flexibility as well as the PSO method.