基于改进FOA的风机叶片结构损伤检测

Fan blade structure damage detection based on improved FOA

为解决风轮叶片多损伤情况下的结构损伤识别问题,提出一种基于单元模态应变能变化率和改进果蝇优化算法的分步损伤判别方法。该方法首先采用单元模态应变能变化率对损伤进行初步判定,然后利用果蝇优化算法对损伤位置和程度进行精确识别。由于传统果蝇优化算法极易陷入局部最优问题,则对其进行了改进。结果表明:采用单元模态应变能变化率可以有效识别出可能的损伤单元,在此基础上用改进果蝇优化算法可以更精确地识别结构损伤位置和程度,同时采用改进果蝇优化算法的识别精度明显优于简单果蝇优化算法。

In order to solve the problem of structure damage identification of wind turbine blade with more damage,a method based on unit mode strain energy difference and improved Fruit fly Optimization Algorithm（FOA） is adopted.Unit mode strain energy difference is used to judge damage preliminarily,then the FOA is adopted to identify the damage location and degree accurately.Because traditional FOA is esay to fall into local optimal problems, so some improvements is done towards it.The results show that unit mode strain energy can judge possible damage unit effectively,then the FOA is adopted can identify structural damage location and degree more accurately, at the same time,the identification accuracy of the improved FOA is superior to traditional FOA.