摘要
为了解决风机叶片损伤类型识别的问题,提出了一种基于谐波小波包和支持向量机相结合的声发射源识别方法。由叶片损伤产生的声发射信号经过4层谐波小波包分解后,提取各频段的能量作为特征向量构建支持向量机分类器,通过支持向量机判别叶片损伤类型。在对叶片损伤进行识别时,分别采用谐波小波包和Daubechies小波包分解声发射信号,并进行比较。实验结果表明,采用谐波小波包和支持向量机相结合的方法可以得到良好的识别效果。
In order to solve the wind turbine blade damage type identification problem,a new approach for acoustic emission （AE） source type identification based on harmonic wavelet packet （HWP）and support vector machine is proposed.The type of blade damage was distinguished by SVM which was built by using the energy as the feature vectors for the support vector machine classifier,where the energy was extracted in different frequency bands from the acoustic emission generated by the blades after a four-level decomposition of HWP.In recognition of the blade for damage,the AE signals were decomposed using harmonic wavelet packet and daubechies wavelet packet and compared with each other.The results show that good recognition results could be obtained using HWP and SVM combined method.
出处
《玻璃钢／复合材料》
CAS
CSCD
北大核心
2014年第4期37-41,共5页
Fiber Reinforced Plastics/Composites
基金
兰州交通大学科技支撑基金资助项目(ZC2012008)
关键词
风机叶片
声发射
谐波小波包
支持向量机
wind turbine blade
acoustic emission
harmonic wavelet packet
support vector machines
