摘要
综述了复合材料损伤失效的声发射检测研究进展,从损伤源定位及损伤模式的识别与分类两方面进行了介绍。在损伤源定位方面模态声发射相比模式识别更有效,在损伤模式的分类方面模式识别技术更加有效,且人工神经网络及小波神经网络在复合材料声发射方面的研究较多。另外,介绍了模糊模式识别技术用于声发射信号分类及聚类的研究情况,根据复合材料声发射信号复杂重叠性的特点,模糊理论结合模式识别技术可以进一步实现复合材料声发射信号更有效的分析。
The development of acoustic emission on damage and failure mechanics of composite materials are briefly reviewed. Both AE damage source location and the classification of damage modes in the previous field are over- viewed. In terms of AE source location, modal acoustic emission technology is more effective than pattern recognition. In terms of the classification of damage modes, pattern recognition is much more effective, and artificial neural net- work and wavelet neural network are more used in acoustic emission on composites. In addition, fuzzy pattern recogni- tion on acoustic emission of classification and clustering of damage modes are discussed. According to the complexity and overlapping of the acoustic emission signal of composites, fuzzy theory together with pattern recognition technology can further make the acoustic emission signal of composite much more effective.
出处
《材料导报》
EI
CAS
CSCD
北大核心
2013年第17期19-22,47,共5页
Materials Reports
关键词
复合材料
声发射
模式识别
神经网络
模糊模式识别
composite, acoustic emission, pattern recognition, neural network, fuzzy pattern recognition