RBFN-based temperature compensation method for impedance monitoring in prestressed tendon anchorage

In this study, a temperature compensation method based on radial basis function network (RBFN) is developed to monitor damage-induced impedance signatures in prestressed tendon anchorages under temperature variation. First, an impedance monitoring technique via a piezoelectric interface device is outlined for prestress-loss detection in the tendon anchorage system. Second, an RBFN-based algorithm is designed to filter out the temperature-induced variation in impedance signatures for damage monitoring. Finally, the feasibility of the proposed algorithm is verified for the prestress-loss monitoring in the test structure under the temperature-varying condition. A series of impedance measurements are conducted for the temperature-varying and prestress force-varying conditions. The RBFN-based algorithm is used to train a set of baseline impedance signatures for varying temperatures, and also to quantify damage-induced impedance signatures of the test structure. The effect of the size of RBFN training patterns on the accuracy of the proposed temperature compensation algorithm is also evaluated.