Damage identification of multi-story steel frames using neural networks

摘要

In this study the basic operations and training method of neural network are introduced and the learning capability of network is explained from the mathematical point of view. The "knowledge" or "memory" can be stored in the parameters of Taylor can be regarded as a general type of NARMA model which is a suitable representation of nonlinear discrete time system. Application of the method to the identification of building seismic response data was performed. The results show that the input-output mappling model from acceleration response is possible for emulating the mdof building structure.