人工神经网络在结构损伤识别中的应用

综合论述了人工神经网络在损伤识别中的作用，主要探讨了分步识别理论和输入参数的选择，并介绍了几种改进的BP算法，最后指出了人工神经网络在高层建筑损伤识别中应用所需要解决的问题。

Application of Artificial Neural Networks to Damage Identification of Structures

It is known that artificial neural network is a promising computing tool for structural damage detection. The application of artificial neural networks to structural damage identification is presented in this paper. For the damage identification of tall buildings, a hierarchical identification approach is introduced, and the selection of input vectors is discussed. Several improved BP algorithms, such as genetic algorithm, dynamic learning rate training, and so on, are introduced. Problems for solution in the application of BP network to damage identification of tall buildings are posed.