基于模拟退火机制的人工蜂群算法的结构损伤识别

发布时间：2018-02-09 07:35

本文关键词： 人工蜂群算法模拟退火算法残余力向量损伤识别　出处：《中山大学学报(自然科学版)》2017年02期 　论文类型：期刊论文

【摘要】：人工蜂群算法是模仿蜜蜂行为提出的一种群智能优化算法。它的主要特点是只需要对问题的解进行优劣的比较,通过各人工蜂个体的局部寻优行为,最终在群体中使全局最优值突现出来,具有较快的收敛速度,但较容易陷入局部最优解。为了克服这一不足,将模拟退火算法机制引入其中进行改进。既保留了蜂群算法群体寻优的特点,又可以有效地避免陷入局部最优解。通过选择合适的收益率函数和温度下降函数,可以很方便地解决优化问题。通过构造基于残余力向量的损伤识别目标函数,利用改进的人工蜂群算法,能有效地解决结构损伤识别问题。通过对桁架模型进行数值模拟,结果表明文中算法就原算法而言,收敛速度,识别精度和抗噪声能力有较好改善。
[Abstract]:Artificial bee colony algorithm is a population intelligent optimization algorithm, which imitates bee behavior. Its main feature is that it only needs to compare the advantages and disadvantages of the solution of the problem, and through the local optimization behavior of each worker bee, In order to overcome this deficiency, the global optimal value will emerge in the population, which has a faster convergence rate, but is more likely to fall into a local optimal solution. The mechanism of simulated annealing algorithm is introduced to improve the algorithm. It not only preserves the characteristics of colony optimization, but also effectively avoids falling into local optimal solution. By selecting the appropriate rate of return function and temperature descent function, By constructing the damage identification objective function based on residual force vector and using the improved artificial bee colony algorithm, the problem of structural damage identification can be effectively solved, and the truss model can be numerically simulated. The results show that the convergence rate, recognition accuracy and anti-noise ability of the original algorithm are improved.
【作者单位】：中山大学工学院;
【基金】：国家自然科学基金(11172333,11272361) 广东省自然科学基金(2015A030313126) 广东省科技计划项目(2014A020218004,2016A020223006)
【分类号】：TP18;O346.5

【相似文献】