基于模态应变能和BP神经网络的混凝土框架结构损伤识别

采用模态应变能变化率和BP神经网络的两步损伤识别方法,对钢筋混凝土框架结构进行损伤识别研究.首先,利用单元的模态应变能变化率作为结构损伤定位因子,运用通用有限元ANSYS的APDL语言编制程序,计算模态应变能变化率;然后,利用BP神经网络进行损伤程度识别.该方法将二者优点相结合,并克服了单纯利用模态应变能难以准确计算损伤程度以及单纯利用BP神经网络样本巨大的困难.数值仿真结果表明,该方法适用于框架结构的损伤识别.

Damage identification for a concrete frame structure by using modal strain energy and BP neural network

Damage identification of reinforced concrete frame structures is investigated by using the changing ratio of modal strain energy and BP neural network.The changing ratio of modal strain energy MSECR is taken as the damage location factor.MSECR is calculated by using the language APDL in the commercial FEM software ANSYS.Then,BP neural network is used to determine the damage degree.Modal strain energy method can not know the damage degree while the BP neural network method has to treat many data of samples.The above shortcomings have been overcome.The results show that the proposed method can be applied to the damage detection of frame structures.