Structural Damage Identification Based On RBF Networks

Structural health monitoring is a hot research subject in civil engineering nowadays. Structural damage identification is the base and center of health monitoring. The technologies of vibration-based analysis and artificial neural networks are the available tools to solve the problem. This paper discusses the theory and application of structural damage identification based on RBF neural networks in civil engineering and the main research and conclusions are summarized as follows:1. It discusses present research and development situation of damage identification in civil engineering and explains the basic theory of damage identification based on modal analysis and neural networks. It introduces artificial neural networks and discusses the algorithms of RBF networks.2. It analyzes the dynamic response of many different damage cases of simple I-shape beam structure and discusses their influences to natural frequency and modal curvature. It is found that the natural frequency and modal curvature can be used in identifying the severity and location of structural damage respectively. Dynamic response of some real beams are measured and compared to the results analyzed by computer, it is found that there are some differences between them and the main reason is the error caused by modeling.3. Based on the dynamic response of all damage cases, it discusses the indexes including frequencies, modal curvatures and their compounded index used in RBF networks. It is found that the method can identify the location and severity of the damage accurately and effectively and the compounded index is better than others.4. Based on the modal analysis of the cable-stayed bridge in many damage cases and the application of damage identification method based on RBF networks, it is found that the damage identification method based on RBF networks can also be applied to the complicated structures such as cable-stayed bridge and it is not more difficult to implement than that of simple beam-like structure.