Damage identification of frame structures based on multi-index curvature modes and neural network

This paper proposes a framework structure damage identification method based on the combination of multi-index curvature mode and neural network. The inherent characteristics and material parameters of the frame are obtained through modal tests and tensile tests. The finite element model is corrected by the full factorial design method in ISIGHT and ABAQUS joint simulation, and the damage data is generated. The central difference method is used to calculate the curvature mode difference (MDC) as the damage identification index, and further proposes the superimposed curvature mode difference (SMDC) and curvature mode difference change rate (CMDCR) indexes. Through the comparison of multi-index curvature modes, the curvature mode difference is successfully applied to the damage identification of frame structures, and the accuracy of damage identification under single-order modal shapes is improved by using multi-indexes.