基于集成智能算法的发动机滑油系统融合诊断

为确保飞行训练安全、预防发动机维修事故发生,针对当前航空发动机磨损故障诊断存在的问题,提出基于集成智能算法的发动机滑油系统故障检测模型.从铁谱诊断、光谱诊断、颗粒计数诊断、理化指标诊断和试车台数据诊断等方面研究了运用集成神经网络方法对多源信息进行融合诊断.结果表明:融合诊断结果的故障模式比单项诊断结果的故障模式要多；当单项诊断出现矛盾时,融合诊断结果能很好地解决诊断冲突问题.基于集成神经网络的发动机油液系统磨损故障的融合诊断能充分利用多种方法的互补性和有效解决诊断冲突问题,从而使诊断结果更为可靠和准确.

Fusion Diagnosis of Engine Lub Oil System Based on Integrated Intelligent Algorithm

In order to ensure the safety of flight training,prevent engine maintenance accident,in view of the current problems of aircraft engine wear fault diagnosis,a fault detection model of the engine lubricating oil system is proposed based on integrated intelligent algorithm.From the ferrographic diagnosis,spectrum diagnosis,particle counting diagnosis,physical and chemical indicators diagnosis,and engine test bed data diagnosis,the diagnosis of multi source information fusion based on integrated neural network method is studied.Results show that the fault mode of fusion diagnosis is more than single diagnosis failure mode.When there is contradiction between individual diagnosis,fusion diagnosis results can well solve conflict problems.The wear fault fusion diagnosis for engine oil system based on integrated neural network can make full use of the complementarity of various methods and effectively solve the problem of diagnosis conflict,so that the diagnosis results are more reliable and accurate.