摘要
在研究Kohonen自组织映射网络理论的基础上运用模糊理论方法建立了刹车系统模糊故障诊断模型。该模型只需选择足够的具有代表性的故障样本训练神经网络，将代表故障的信息输入给训练好的神经网络，根据神经网络的输出结果，就可以判断发生故障的类型。该模型除能识别已训练过的故障，还能识别未训练过的故障，并且聚类能力强、速度快，因此很符合复杂系统的故障诊断。
According to features of aircraft brake system failure and its failure type and the uncertainty for its fault knowledge representation, the paper studies the principles of Kohonen self-organizing map neural network and fuzzy method of fault diagnosis, and builds the fault diagnosis model on fuzzy neural network. It need not to set up a complicated mathematical model, and also need not to do complicated mathematical calculation and data processing. All you need do to is select enough typical fault samples to trained neural network, the type of shown fault can be judged by the output of the neural network. It cant only discriminate faults for which it has been trained, but also such for which it hasnt been trained. It has a strong grouping capability, works quickly and is therefore suitable for fault diagnoses of complicated systems.
出处
《计算机工程》
CAS
CSCD
北大核心
2003年第14期98-99,共2页
Computer Engineering
基金
中国民用航空总局教育研究基金资助项目(02-3-11)
关键词
自组织映射
故障诊断
神经网络
Self-organizing map
Fault diagnosis
Neural network
