水电数字仿真在线融合故障诊断专家系统

Expert System of Fault Diagnosis Based on On-line Fusion in Hydroelectricity Digital Simulation

为了解决丰满水电数字仿真系统的在线故障诊断问题，基于信息与方法融合的思想 ，提出了分布式集成神经网络建模方法、模糊神经网络专家系统(FNNES)在线故障诊断方法 。将模糊神经网络(FNN)嵌入专家系统(ES)中，FNN负责知识获取和逻辑推理，ES负责系统信 息的输入和输出、符号推理，并对FNN的结论进行解释。系统的运行验证了方法的有效性和 实际应用价值。为现场诊断系统的开发提供了有益的方法和经验。

In order to solve the problem of on-line fault diagnosis in Fengman hydropower digital simulation system, a distributed integrated neural network m odeling method, a on-line fault diagnosis method of fuzzy neural network expert system(FNNES) are proposed based on information and method fusion. The fuzzy neu ral network(FNN) is embedded to expert system(ES), FNN takes charge knowledge ac quisition and logic inference, ES takes charge input/output of system informatio n and symbolic inference and explain conclusion of FNN. FNNES has been tested an d run, it shows that the proposed method is effective, validated its generality, advantage and practical application value. The beneficial methods and experienc es are provided to the development of locale fault diagnosis system.