核动力系统神经网络故障诊断专家系统研究

以核动力系统故障诊断专家系统所存储的产生式规则库为依据，构造专家系统人工神经网络模块，建立具有并行推理能力的神经网络推理机制和实例学习能力的知识获取机制。在此基础上，建立了针对所构建神经网络的推理解释机制。系统融合了传统专家系统和神经网络各自优势，能高效、准确地进行实例推理、诊断解释并能有效地从实例中获取知识。

Research on Neural Network Fault Diagnosis Expert System of Nuclear Power System

According to the production rules base stored in the nuclear power system fault diagnosis expert system, a neural network (NN) module of expert system was built. The NN inference mechanism including the ability of parallel inference and the knowledge acquisition mechanism was set up. On the basic of the built NN, an inference explaining mechanism was built to it. This system syncretizes advantages of both traditional expert system and NN that it can do case inference and diagnosis, and explain efficiently and accurately as well as effectively acquire knowledge from real cases.