基于模糊神经网络的水轮机调速器故障诊断

Fault diagnosis of hydro turbine governor based on fuzzy neural networks

分析了水轮机调速器调节过程的故障机理,得到专家经验的故障诊断推理规则,利用模糊逻辑在表达专家知识方面的优势及神经网络的自学习能力,建立模糊神经网络,利用有限的推理规则对模糊神经网络进行训练,得到的模糊神经网络作为专家系统用于水轮机调速器的故障诊断,解决了故障样本获取困难、专家经验获取不足及模糊规则"组合爆炸"的问题,仿真实验证明该网络诊断结果正确,实用性较强。

This paper analyzes fault mechanism of regulating process for hydraulic turbine governor and establishes rules of fault diagnosis according to experts′ experiences.A fuzzy neural network is developed by taking advantages of expert knowledge expression and its self-learning ability.It can be trained with a number of diagnosis rules.As an expert system,it was applied to fault diagnosis for a hydraulic turbine governor,overcoming the difficulties in sample collection,insufficient expert experiences and ′combination explosion′problem of fuzzy rules.Simulation results show that it is a feasible and effective method in fault diagnosis.