延性节点钢框架结构损伤多重分步识别仿真

Multi-step Damage Identification Simulation of Steel Frame Structure With Ductile Joints

延性节点钢框架结构损伤会对建筑整体稳定性和抗震性造成严重影响,因此保证延性节点钢框架结构一直处于健康状态具有重要的现实意义。针对主成分分析方法、基于统计矩方法和模态参数识别方法等三种传统结构损伤识别方法准确性低的问题,研究一种新的延性节点钢框架结构损伤多重分步识别方法。方法首先利用无线传感器系统采集延性节点钢框架结构的振动响应数据,然后利用主成分分析方法对响应数据进行降维处理,再然后利用小波分析方法定位结构损伤位置,最后利用人工免疫算法确定钢框架结构损伤程度。结果表明,与三种传统结构损伤识别方法相比,利用本方法进行延性节点钢框架结构损伤识别,F分数值更接近1,由此证明本方法识别损伤的准确性更高。

The damage of steel-frame structure with ductile connection will seriously affect the overall stability and seismic resistance of the building,so it is significant to ensure that the steel-frame structure with ductile connec⁃tion is always healthy.Due to low accuracy of principal component analysis method,statistical moment method and modal parameter identification method,a method of multi-step identification for steel-frame structure damage with ductile connection was studied.At first,this method used the wireless sensor system to collect vibration response data of steel-frame structure with ductile connection.And then,the principal component analysis method was used to re⁃duce the dimensionality of response data.After that,the wavelet analysis method was used to locate the damage loca⁃tion.Finally,the artificial immune algorithm was adopted to determine the extent of structural damage.Compared with traditional methods,the F-score value obtained by the proposed method is closer to 1.Therefore,the proposed meth⁃od is more accurate in identifying damage.