基于改进混合蛙跳算法的结构多缺陷反演

Detection of Multiple Flaws Based on Improved Hybrid Leapfrog Algorithm

将改进混合蛙跳算法与动力扩展有限元法相结合,建立了基于结构振动响应的结构多缺陷反演模型.在混合蛙跳算法中,结合花授粉算法中的转换概率,在更新青蛙个体位置时引入子群内随机个体进行协助搜索,提高算法跳出局部最优的能力;同时引入双中心策略,提高优化算法解的质量.将改进混合蛙跳算法作为缺陷信息反演分析算法,动力扩展有限元法作为数值分析方法计算结构响应,结合二者建立结构多缺陷反演模型.将该模型应用于结构孔洞、裂纹等多种缺陷的反演分析,研究该算法在高斯噪声影响下的鲁棒性.结果表明:结构多缺陷反演模型能够快速准确反演出结构多种复杂缺陷信息,改进混合蛙跳算法相比于经典混合蛙跳算法具有收敛速度快,结果精确度高,鲁棒性强的特点.

In this paper, the improved shuffled frog leaping algorithm(SFLA) is combined with the dynamic extended finite element method to establish a structural multiple defect inversion model based on structural vibration response. In the SFLA, combined with the conversion probability in the Flower Pollination Algorithm, the random individuals in the subgroup are introduced to assist the search when updating the individual position of frogs, so as to improve the ability of the algorithm to jump out of the local optimum. The double center strategy is introduced to improve the quality of the solution of the optimization algorithm. The improved SFLA is used as the defect information inversion analysis algorithm, and the dynamic extended finite element method is used as the numerical analysis method to calculate the structural response. Combined with two algorithms, the structural multi defect inversion model is established. The model is applied to the inverse analysis of structural defects such as holes and cracks, and the robustness of the algorithm is studied. The results show that the structural multi defect inversion model can quickly and accurately reflect a variety of complex defect information of the structure. Compared with the classical SFLA, the improved SFLA has the characteristics of fast convergence speed, high accuracy and strong robustness.