声发射信号具有时变的特征,对声发射信号进行短时分析技术,并采用快速傅里叶变换后提取信号的有效特征.进而采用fisher准则对信号特征数量进行压缩,最后采用神经网络技术对信号进行分类,分类结果表明这种方法对声发射信号特别有效。
In view of the time-varying characteristics of AE(Acoustic Emission) signals, the short-time analysis and FFT(Fast Fourier Transform) are used to extract the signal characteristics of effective. Then the fisher criterion is used to compress the number of signal. Finally the neural network is employed for signal classification. Classification results show that the method is particularly effective in terms of the AE signal.