基于混沌人工蜂群算法的结构损伤识别

Damage Identification Based on Chaotic ABC Algorithm

人工蜂群算法作为一种模拟蜜蜂采蜜行为提出的优化算法,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解突显出来,具有较好的收敛速度,有广泛的工程应用前景。基于人工蜂群算法和MAC(模态保证准则)对强、弱耦合双梁系统进行局部损伤的识别。为了避免算法过早陷入局部最优解,引入锦标赛机制和混沌搜索机制对算法进行改进。算例表明此方法能有效快速的检测出耦合双梁系统的局部损伤、对测量噪声不敏感,较原算法相比,具有更高效率、更高精度等优点,有望用于实际工程应用。

Artificial bee colony（ ABC） algorithm is an effective approach to solve structural local damage detection. The damage detection in a double-beam system coupled via a set of springs based on a modified artificial bee colony algorithm is discussed. By using the system natural frequencies and modal assurance criteria（ MAC）,the ABC algorithm is utilized to solve the nonlinear objective function from the coupled double-beam system. To avoid the algorithm trapped local optimize easily,the ABC algorithm is modified by choosing the tournament selection instead of roulette,and chaotic search mechanism is applied to improve algorithm＇s global search capability. The simulation results shows that the modified ABC algorithm can identify the local damage of the coupled system effective even under measurement noise,better than the original algorithm.