应用神经网络信息融合诊断航空发动机故障

Fault diagnosis for aeroengine gas path components based on neural network multisensor data fusion

研究了基于神经网络信息融合技术,同时结合模糊集合论对发动机气路部件进行故障诊断的方法,并以某型涡轴发动机为对象进行了仿真分析.研究结果表明该方法的故障诊断过程相对简单,对模型的精度要求不高,能够降低虚警、误报、漏报等情况的发生.

A fault diagnosis method for aeroengine gas path components, which was based on the technology of neural network multisensor data fusion and combined application of fuzzy logical theory, was studied in the paper. And this method was employed in simulation analysis of a given turbo-shaft engine. The whole diagnosis process of this method is simple and high accuracy of modeling is not required. Fault alarm, mistaken report and false positive can be reduced.