APPLICATION AND ANALYSIS OF DIFFERENT INFORMATION FUSION METHODS FOR STRUCTURAL DAMAGE DETECTION

Different information fusion approaches possess different sensitivity and computational complexity in structural damage detection field.In order to solve the problem,a functional fusion model for damage identification is presented.Three main fusion methods,i.e.Bayesian fusion method,Dempster-Shafer evidence theory and fuzzy set theory,which are based on the functional fusion model,are proposed to identify the structural multiple damage sites.Simulation results demonstrate the excellent performance of information fusion technique to identify damage locations as compared with the methods based on single information source.The applicability scope of different fusion approaches is pointed out.