Multi-stage Damage Identification for Frame Structures Based on Neural Network

The multi-stage damage identification method for frame structures based on neural network was proposed.A kind of high efficient neural network to identify damage in frame structures was established.This method was divided into four steps according to the multi-stage identification ideas of member damage.Firstly,damage anomalous filter was set up by neural network to alarm the damage in structural members.Secondly,the primary location of the member damage was determined by the neural network with inputing the combined damage index of frequency.At the third step,the specific location of the member damage was determined by the neural network with inputing the combined damage index of frequency and vibration mode.Finally,the damage degree of the member was identified by neural network with inputing the change rate of squared modal frequency.Numerical simulation of damage identification for three-span four-floor frame structure was carried out.Results show that multi-stage damage identification method based on neural network simplifies network structure,and can alarm,locate and quantify the damage effectively.