论文部分内容阅读
提出了一种基于经验模态分析(Empirical mode decomposition,EMD)和D-S证据相结合的飞行器健康诊断方法。该方法首先对由声发射传感器募集到的飞行器关键结构部件原始声发射信号进行EMD,得到多个内禀模态分量,选取内禀模态能量构建声发射信号的特征向量,并分别采用模糊神经网络、GRNN网络和Elman神经网络对提取出的特征向量进行分类,最后运用D-S证据理论进行决策融合,对飞行器的健康状态进行诊断。实验表明,运用此方法对某型号真实飞行器关键结构部件的健康状态进行诊断,可以得到很好放入诊断结果。与单分类器相比,采用D-S证据理论进行决策融合有效地提高了故障诊断的精度。
A method of aircraft health diagnosis based on empirical mode decomposition (EMD) and D-S evidence is proposed. In this method, the original acoustic emission signals of the key structural parts of the aircraft recruited by the acoustic emission sensors are EMD, and a plurality of intrinsic mode components are obtained. The intrinsic mode energy is selected to construct the eigenvectors of the acoustic emission signals, and the fuzzy neural Network, GRNN network and Elman neural network to classify the extracted eigenvectors. Finally, DS evidence theory is used to make decision fusion to diagnose the health of the aircraft. Experiments show that using this method to diagnose the health status of the key structural components of a real aircraft can be put into the diagnosis well. Compared with single classifier, using D-S evidence theory to make decision fusion effectively improves the accuracy of fault diagnosis.