A Global Artificial Fish Swarm Algorithm for Structural Damage Detection

Abstract

Structural damage detection (SDD) is an important but still challenging task in the structural health monitoring (SHM) field. Many methodologies have been developed and broad application prospect are expected. However, there are still some difficulties when they are applied to the real structures. In this study, a novel global artificial fish swarm algorithm (GAFSA) is proposed for exploring a new solution to the SDD problem in the SHM field. Firstly, the basic theory of the GAFSA is introduced. The fish swarm behaviours inside water are simulated by the following four steps: random, preying, swarming and following behaviours, respectively. The artificial fish parameters are defined, the implementing procedure of GAFSA is expressed, and the computing performance of GAFSA is evaluated and compared with the basic artificial fish swarm algorithm by three test functions. Secondly, the SDD problem is modelled as a constrained optimization problem in mathematics, an objective function on optimization problem is defined, and the model updating-based SDD is hopefully solved by the proposed GAFSA, which is based on swarm intelligence and uses a population (or swarm) of fish to identify promising regions looking for a global solution. Some numerical simulations on single and multiple damage cases of both an ASCE 4-storey benchmark frame structure and a 2-storey rigid frame have been conducted for assessing the effectiveness and robustness of the proposed GAFSA. Finally, a laboratory experimental study on damage detection of a 3-storey building model with four damage patterns was performed. The illustrated results show that the proposed GAFSA can not only locate the structural damage but also quantify the severity of damage with a good noise immunity.