Inverse Analysis for Identification of a Truss Structure with Incomplete Vibration Strain

摘要

The increasing use of advanced sensing technologies such as optic fiber Bragg grating and embedded piezoelectric sensors necessitates the development of strain-based identification methodologies.Recently,a three-step neural networks based structural inverse analysis strategy,called direct soft parametric identification(DSPI),with a strain-based emulator neural network(SENN)and a parametric evaluation neural network(PENN)has been presented to identify structural stiffness and damping parameters directly from free vibration-induced strain measurements using an evaluation index called root mean square of prediction difference vector(RMSPDV).In reality,because the number of strain sensors is limited and it is difficult to get strain information for all members of a large-scale structure,it is necessary to study the performance of the proposed methodology when incomplete measurements are available.The performance of the proposed methodology using spatially incomplete vibration strain measurements is examined by numerical simulations with a truss structure involving all stiffness values unknown.Numerical simulation results show that the proposed methodology is a practical method for near real-time identification and damage detection with spatially incomplete vibration-induced strain measurements.