遗传算法在结构损伤模式识别中的应用  

Application of Genetic Algorithms in the Pattern Recognition of the Damage Structure

作　　者：王广伟[1] 李强 徐海峰 

机构地区：[1]浙江海洋学院数理与信息学院,浙江舟山316000 [2]浙江省舟山市岱山县大衢中学计算机教研室,浙江岱山316181 [3]浙江省舟山市舟山航海初级中学,浙江舟山316000

出　　处：《浙江海洋学院学报（自然科学版）》2008年第3期335-339,共5页Journal of Zhejiang Ocean University(Natural Science Edition)

基　　金：浙江海洋学院科研启动基金

摘　　要：研究了遗传算法在结构损伤诊断中的应用。将结构的损伤诊断问题等价为结构的损伤模式的识别问题,并应用遗传算法的全局搜索功能来实现损伤状态的优化识别。其中应用结构的第一阶振型变化率作为结构的损伤诊断标识量,从而减小遗传操作的种群规模,极大的提高了遗传算法的计算效率。四边固支板结构的单位置以及多位置损伤诊断算例说明了该方法是可行的,具有较高的识别精度。In the paper, Genetic Algorithms is studied to detect the damage of the structure. The new method transforms the damage detection of structure into the recognition of the damage patterns, and accomplishes the recognition with the global search of Genetic Algorithms. During the process, with the variety rate of the first modal as the signature for damage detection, the scale of the tribe is reduced and the efficiency is improved. The examples of the single damage detection and multiply damages detection of the rectangle plate fixed at four sides show that the method is not only feasible, but also with high precision.

关 键 词：损伤标识量 遗传算法 损伤诊断 模式识别 全局搜索 

分 类 号：O327[理学—一般力学与力学基础]