论文部分内容阅读
基于递阶遗传算法(HGA)与结构优化思想,提出了一种针对欧拉-伯努利梁和二维板结构的多损伤监测方法.该方法利用递阶遗传算法的控制基因表示损伤的数量和位置,以参数基因表示损伤的程度,有效地避免了传统遗传算法(CGA)的早熟现象所造成的损伤误识别等问题.一个悬臂梁和悬臂方板结构模型的多损伤监测仿真计算表明该方法能够准确地监测一、二维结构中多个位置的损伤,而传统遗传算法难以识别二维结构中的多损伤情况.悬臂梁仿真算例中,该方法和传统遗传算法对多损伤程度的识别误差分别为0.144%和1.819%,所需的有限元计算次数该方法仅为传统遗传算法的16.4%.与传统遗传算法相比,递阶遗传算法明显提高了损伤识别方法的计算效率、精度和稳定性.
Based on hierarchical genetic algorithm (HGA) and structural optimization, a multi-damage monitoring method based on Eulerian-Bernoulli beam and two-dimensional plate structure is proposed, which uses the control gene of a hierarchical genetic algorithm to represent the number of damage And the position of the cantilever beam to indicate the extent of the damage by the parameter gene effectively avoids the mistaken identification of the damage caused by the premature phenomenon of the traditional genetic algorithm (CGA), etc. The simulation calculation of multiple damage monitoring of a cantilever beam and a cantilever square plate structural model indicates that the The method can accurately monitor the damage in one or two-dimensional structure in many places, but the traditional genetic algorithm can not identify the multi-damage in the two-dimensional structure.In the cantilever simulation example, this method and the traditional genetic algorithm have much more damage degree The recognition errors are 0.144% and 1.819%, respectively, and the required number of finite element calculations is only 16.4% of the traditional genetic algorithm.Compared with the traditional genetic algorithm, the hierarchical genetic algorithm obviously improves the computational efficiency and accuracy of the damage identification method And stability.