基于神经网络的空间桁架结构损伤三重识别

以一72杆空间钢桁架为例进行桁架结构损伤三重识别．考虑单损伤工况与双损伤工况，通过ANSYS软件建模得到结构在无损与损伤工况下的固有频率，运用神经网络进行结构损伤定位与损伤程度确定．首先识别损伤层，然后识别损伤层中的损伤杆件，最后识别损伤杆件的损伤程度．结果表明该方法用于桁架结构损伤识别是可行的．

A Three--step Damage Identification Approach Based on Neural Network for Space Truss Structure

A three--step damage identification approach is carried out on a 72--truss space structure. The model is built up by ANSYS in order to get the natural frequencies in healthy and damage conditions. The single--damage condition and double--damage condition are considered. Neural network is adopted to determine the location and the extent of damage. The first step is to determine the damage layer, then to determine the damage member in the damage layer, and the last step is to determine the extent of damage. The results illustrate that the three--step damage identification approach is efficient in damage identification of truss structure .