摘要
介绍了燃料电池发动机的常见易发性故障,结合实际调试经验和燃料电池发动机的工作机理,在收集大量故障样本数据的基础上,针对其执行器故障和误用故障提出了一种基于模糊神经网络的故障诊断方法并建立其故障诊断模型.该方法利用模糊逻辑理论将输入量和输出量均以隶属度表示,然后通过故障样本对人工神经网络进行训练,从而实现从故障征兆到故障原因的推理.实际应用和实验仿真结果表明该方法推理速度快、容错能力强,建立的故障诊断模型具有较好的实用性.
The common malfunctions of fuel cell engine(FCE) was discussed in this paper,according to the actual debugging experience and working mechanism of FCE,on the basis of large quantities of collected faults sample data,a kind of fault diagnosis method based on fuzzy neural networks(FNN) which was applied in FCE′s actuator faults and misuse faults was put forward and the fault diagnosis model was set up as well.To carry out the method and realize the inference from fault symptoms to fault reasons,fuzzy logic theory was used to make inputs and outputs expressed with membership degree,and then the faults sample was employed to get artificial neural networks(ANN) trained.The results of practical employment and experimental simulation prove that the method adopted has good ability of quick reasoning speed and robust fault-tolerance,and the fault dignosis model setted up is practicable.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2009年第S1期114-117,139,共5页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家高技术研究发展计划资助项目(2006AA11A133)
