摘要
以数据融合技术进行桁架结构的单损伤和多损伤识别。通过研究基于频率的结构损伤理论,分析归一化的频率和损伤位置的关系;利用小波概率神经网络的算法对决策融合进行修正,建立基于小波概率神经网络的数据融合结构损伤识别模型。运用结构计算软件计算了一典型桁架结构的频率,并融合为小波概率神经网络算法的输入特征向量,并对桁架算例模型结构进行损伤识别。通过桁架不同位置的损伤情况,验证该方法的有效性,并提出工程应用中应注意的问题。研究结果表明,基于小波概率神经网络算法的数据融合技术是一种比较可靠的损伤识别方法,具有良好的工程应用前景。
Data fusion techniques were used to do single-damage and multi-damage identification of truss structures. By studying the structural damage theory based on frequency, it was analysed the relation between the normalized frequency and location of damage; and the wavelet probabilistic neural network algorithm was used to modify decision fusion, thus establishing a data fusion model of structural damage identification based on wavelet probabilistic neural network. The structural calculation software was used to calculate the frequency of a typical truss structure, which was integrated into an input feature vector of the wavelet probabilistic neural network algorithm, and a damage identification was done for the truss structure model example. Through truss damage of different locations to verify the effectiveness of the method was verified , and the problems needing attention during engineering uses was proposed. The results showed that data fusion technology based on wavelet probabilistic neural network algorithm was a more reliable damage identification method, and had good prospects for engineering applications.
出处
《工业建筑》
CSCD
北大核心
2012年第12期129-132,共4页
Industrial Construction
关键词
结构损伤
桁架结构
损伤识别
数据融合
小波概率神经网络
structural damage
truss structure
damage identification
data fusion
wavelet probabilistic neural network
