【摘要】 对于复杂的诊断对象 ,本文提出了一种复合模糊神经网络结构 ,该神经网络结构集成了一系列模糊神经子网络 ,来完成故障分类任务。文中以某高温硝酸冷却系统为诊断对象 ,以系统故障树为设置子网的基础 ,建立了含有多个子网的复合模糊神经网络。仿真结果表明了这种方法的可行性和有效性更多还原

【Abstract】 In this paper, a hybrid fuzzy neural network architecture is proposed for complex diagnosis objects. A series of fuzzy neural sub networks are integrated to perform the task of fault classification. Basing the setting of the sub networks on the fault tree of the system, we build the hybrid neural network that includes several sub networks for the cooling system of the high temperature nitric acid. The simulations show the feasibility and validity of the architecture.更多还原