A statistical method for health diagnosis of concrete bridge based on EWMA control chart and reliability analysis

Health diagnosis of bridge by traditional model-based method would face strong difficulty in constructing an accurately finite element model (FEM) due to the document lack of old bridge and the change of the structural parameters of bridge in the service. The further challenge still lies in the uncertainty of the ambient excitation in structural health monitoring of bridge. In this paper, an improved statistical method without FEM is presented to achieve the health diagnosis of concrete bridge by statistical process control associated with reliability analysis. Firstly the characteristic of the live-load deflection effect of the main girder in a bridge is analyzed as a theoretical basis of the proposed strategy. Then wavelet analysis is employed to extract the live-load effect of deflection from the total measurement. Finally the EWMA control chart is designed to complete the structural health diagnosis of the bridge, in which the multi-stage control limits are estimated by reliability analysis. The results from application in a real bridge show the effectiveness of the proposed method in health diagnosis of bridges.