摘要
油样分析方法目前已成为航空发动机磨损故障诊断的重要手段,但单一油样分析技术的诊断准确率均有限,为了提高故障诊断的精度,本文提出了基于D-S证据理论的发动机磨损故障智能融合诊断方法。首先用BP神经网络实现发动机磨损故障的单项智能诊断,然后,充分利用神经网络诊断结果,用D-S证据理论实现了磨损故障的融合诊断。最后,算例验证了本文方法的有效性。
Oil analysis technology has become a common technology in the filed of aero-engine wear fault diagnosis. The effectiveness of individual oil analysis technology, however, is limited in its accuracy. An intelligent fusion technique based on the Dempster-Shafer（D-S） evidence theory is proposed to improve the diagnosis accuracy. Firstly, the BP neural network is employed to carry out single aspect diagnosis; then the final conclusions are reached by combining the results of different diagnostic tools based on the Dempster-Shafer evidence theory. Examples show the validity of the technique proposed in this paper.
出处
《机械科学与技术》
CSCD
北大核心
2005年第9期1018-1021,共4页
Mechanical Science and Technology for Aerospace Engineering
