摘要
推理技术是故障诊断系统的核心,目前针对复杂系统的故障推理仍然是个难题,提出了将模糊神经网络与事例推理相融合的集成诊断模型,通过规则推理、事例推理和神经网络推理三个子系统的互相验证,既解决了事例推理不能得到最优解的问题,又解决了模糊神经网络训练样本难以选取的问题,将该推理方法用于某武器系统故障诊断,取得了较好地诊断效果。
Inference technology is the most important in fault diagnosis system. However it is difficult to draw an inference in complexity system. The model integrated RBR, CBR and FNN is put forward to ,solve the trouble of armament. The problem that inconvenience to get optimization in CBR and select training-sample in FNN is raveled. Emulator shows that the model has more advantage.
出处
《计算机测量与控制》
CSCD
2007年第2期214-215,221,共3页
Computer Measurement &Control
基金
装备研究院创新工程资助项目(05X7K3101)
关键词
模糊神经网络
复杂系统
故障诊断
fuzzy neural network
complexity system
fault diagnosis
