基于残余力向量法和树种算法的结构损伤识别二步法

发布时间：2018-06-28 07:51

  本文选题：树种算法 + 残余力　； 参考：《中山大学学报(自然科学版)》2017年04期

【摘要】：基于二步法对结构进行了损伤识别。首先,通过改进残余力向量定位方法对结构单元进行排查、找出疑似单元;然后基于频域参数建立目标函数,并利用元启发式树种算法(Tree-seed algorithm)对于结构损伤进行反演。为了改善算法的优化能力,引入了两种局部搜索模式,并采用具有自适应性的判断参数来判定使用何种模式进行局部开发。最后,采用梁和悬臂板作为算例进行损伤识别,并得到了识别结果。结果表明:在仅知道有限频域参数的情况下,提出的算法能够有效地识别损伤参数,优于其他进化算法并且对测量噪声不敏感。
[Abstract]:The damage identification of the structure is carried out based on the two-step method. Firstly, the structure unit is searched by improved residual force vector location method to find the suspected element, and then the objective function is established based on the frequency domain parameters, and the damage of the structure is retrieved by using the Tree-seed algorithm. In order to improve the optimization ability of the algorithm, two local search modes are introduced, and the adaptive judgment parameters are used to determine which pattern is used for local development. Finally, the beam and cantilever plate are used as examples to identify the damage, and the recognition results are obtained. The results show that the proposed algorithm is better than other evolutionary algorithms in identifying damage parameters and is insensitive to measurement noise under the condition of only knowing the parameters in finite frequency domain.
【作者单位】： 中山大学工学院;
【基金】：国家自然科学基金(1117233) 广东省自然科学基金(2015A030313126) 广东省科技计划项目(2014A020218004,2016A020223006)
【分类号】：O346.5
，

本文编号：2077275