【摘要】 本文针对ASCE提出的健康监测基准问题的结构模型,分析了基于固有频率的结构损伤识别方法,使用MATLAB软件模拟不同的结构损伤情况,根据得到的固有频率构造BP神经网络输入参数并进行相应的训练,通过神经网络对结构损伤分步分层次的进行诊断。通过数值分析结果说明这种方法可以取得较好结果,也进一步说明神经网络技术是进行结构损伤识别有效方法。更多还原

【Abstract】 In this paper,with ASCE structural health monitoring benchmark model,a damage assessmentmethod based on natural frequencyis presented,and a step by step process of assessment with the use of BP neural network is analyzed by utilizing MATLAB software,which the network is formed by frequencies of the structure with different damage.The results of the simulated data revealthe method is feasibleand BP neural networkis efficienttechnology for damage assessment of structure.更多还原