摘要
结构损伤诊断子系统是建立结构智能健康监测专家系统的核心问题。采用人工神经网络技术可以实现结构损伤的自动识别与定位,具有广阔的应用前景。介绍了基于人工神经网络的两级损伤识别策略,并对采用人工神经网络进行结构损伤诊断的网络输入参数与网络结构选择等关键问题进行了探讨。
The sub-system of structural health diagnosis is key part of the intelligent health monitoring expert system for large-scale structures. The occurrence and location of the damage may be indicated automatically by the technology of damage identification with artificial neural network, so the technology has bright prospect. The two-step identification strategy based on neural network is introduced. At the same time, the choices of input parameters for networks and of the network constructions for networks that are the important problems in structural damage diagnosis with artificial neural network have been discussed.
出处
《无损检测》
2004年第8期383-387,共5页
Nondestructive Testing
基金
国家重点工程科研项目
关键词
健康监测
损伤诊断
神经网络
动力特性
概率神经网络
Health monitoring
Damage diagnosis
Artificial neural network
Dynamical characteristic
Probabilistic neural network
