论文部分内容阅读
提出了一种基于知识与模糊神经网络专家系统故障诊断方法 ,设计构造了诊断专家系统的整体框架 ,在框架中体现了在任务调度机控制下 ,规则符号推理和模糊神经网络推理综合诊断的思想。知识库是诊断专家系统的核心 ,在本系统中定义了广义三层规则库结构 ,即元规则、故障诊断规则、结论合并规则。 Rule型的模糊联想记忆器实现专家系统中的分类和综合功能 ,并讨论了模糊神经网络输入和输出模糊化的问题。本文为旋转机械故障诊断专家系统提供了一个易于实现的框架结构。以汽轮机故障诊断为例进行了实验分析 ,说明此方法具有推理效率及诊断准确性高的特点
A fault diagnosis method based on knowledge and fuzzy neural network is proposed. The whole frame of diagnosis expert system is designed and constructed. The idea of rule symbol reasoning and fuzzy neural network inference under the control of task scheduling machine . Knowledge base is the core of diagnosis expert system. The generalized three-layer rule base structure is defined in this system, namely meta-rules, fault diagnosis rules and conclusion consolidation rules. The Rule-type fuzzy associative memory realizes the classification and synthesis of expert systems, and discusses the fuzzification of input and output of fuzzy neural networks. This paper provides an easy-to-implement frame structure for rotating machinery fault diagnosis expert system. Taking the fault diagnosis of steam turbine as an example, the experimental analysis is carried out, which shows that this method has the characteristics of reasoning efficiency and high diagnostic accuracy