摘要
给出了将小波奇异性检测原理应用于结构损伤检测的方法:对结构的模态振型进行离散小波变换,根据小波变换模极大诊断结构的损伤位置;利用BP神经网络模拟多个尺度下小波变换模极大与损伤程度之间的非线性关系,根据网络的输出诊断结构的损伤程度.为了检验该方法的有效性,以某简支梁损伤检测为例进行了数值模拟,结果表明,利用离散小波进行结构损伤检测,无需计算Lipschitz指数,而且精度可满足工程要求.
An introduction was given to a structural damage detection method based on the singularity detection principle. In the method, the discrete wavelet deformation was performed of the vibration mode of structures, and the location of structural damage was determined by the modulus maximum; the nonlinear relationship between the modulus maximum of the wavelet transform for different scales and the degree of structural damage was simulated by BP neural network, and the degree of structural damage was determined according to the output of the network. Numerical simulation of damage detection of a simply supported beam demonstrates that, without need for calculation of Lipschitz index, the method is effective for structural damage detection, and the precision meets the requirement of engineering.
出处
《河海大学学报（自然科学版）》
CAS
CSCD
北大核心
2006年第3期302-305,共4页
Journal of Hohai University(Natural Sciences)
基金
国家自然科学基金资助项目(50379005)
水利部科技创新基金资助项目(SCX2000-56)
