Neural Network-Based Damage Detection from Transfer Function Changes

This article puts forth the work on a neural network-based approach to determine the degree of damaged floors of the building considering changes in the transfer function. The transfer function is considered for that part of forced vibration in which system vibrates linearly after the structure has been damaged considering the building is instrumented during the earthquake. The results showed that accuracy of degree of damage detected increased with the increase in the number of combination of damages. The instrumentation of the first floor is expected to give best results for damage detection based on the transfer function-based approach.