飞机疲劳试验裂纹声发射信号的人工神经网络模式识别
The Neural Network Pattern Recognition of Crack AE Signal During Aircraft Fatigue Testing
-
摘要: 对某型飞机疲劳试验过程中采集的信号, 先后采用频谱、多分辨率小波方法对原始信号作了分析, 但这两种方法的分析结果不甚理想, 从而改用人工神经网络模式识别的方法提取出了可能是裂纹信号的波形信号。该试验为类似疲劳试验等强噪声环境下的声发射信号模式识别提供了一种可供参考的方法。Abstract: Utilizing the FFT and wavelet methods to analyze the original crack AE(Acoustic Emission) signal during aircraft fatigue testing, but the results were not ideal. Then, the neural network pattern recognition method was hence applied, and the crack AE waveform was extracted. The experiment put forward a reference to AE signal pattern recognition under strong background noise as fatigue testing.