基于小波分析和遗传算法的结构损伤诊断  

Damage identification of structures based on wavelet analysis and genetic algorithm

作　　者：管德清[1] 于兴虎[1] 陈豫洲 

机构地区：[1]长沙理工大学土木与建筑学院,湖南长沙410004

出　　处：《长沙理工大学学报（自然科学版）》2014年第2期43-47,共5页Journal of Changsha University of Science and Technology:Natural Science

基　　金：国家自然科学基金资助项目(51378079)

摘　　要：提出了小波-遗传算法的概念,建立了一种既能识别结构损伤位置、又能确定损伤程度的小波-遗传算法。首先,以有限元分析求解损伤结构振型模态为基础,用db1小波做连续小波变换,由小波系数模极大值识别损伤的位置。然后,以单元刚度的折减系数为遗传算法的优化变量,用振型和频率的误差函数加权来构造目标函数,并通过损伤位置的确定来简化目标函数的变量,再用遗传算法对目标函数进行优化,从而确定结构的损伤程度。通过对一简支梁进行数值模拟分析,计算结果表明,提出的方法不仅能够有效识别损伤的位置,而且能够准确识别损伤程度。The concept of wavelet and genetic algorithm is put forward.An identification method of wavelet genetic algorithm which can identify both damage location and damage degree of structure is established.First,based on the damaged structure of vibration mode with the finite element analysis,using db1 wavelet to make the continuous wavelet trans-form,identify the damage location by wavelet series maximum.Then the reduction factor of element stiffness is put as optimization variables of genetic algorithm.Using the error func-tion of vibration mode and frequency to construct the obj ective function,the variables of the obj ective function are simplified through the determination of damage location.By optimi-zing the obj ective function with a genetic algorithm,the damage degree of structure can be determined.Through the numerical simulation analysis to a simply supported beam,the calculation results show that the proposed method not only can identify the damage location effectively but also can identify the extent of the inj ury accurately.

关 键 词：小波分析 遗传算法 小波-遗传算法 损伤位置 损伤程度 

分 类 号：O346.5[理学—固体力学] TN911.6[理学—力学]