基于模糊神经网络的燃气轮机故障诊断专家系统研究

针对神经网络和传统专家系统在燃气轮机故障诊断过程中各自存在的局限性,提出了一种将模糊神经网络和专家系统相结合的方法.解决了以往专家系统专家知识获取困难和不能描述模糊性知识的缺陷.通过已开发的某型三轴燃气轮机运行模拟器取得典型的故障样本完成了对模糊神经网络的训练工作,最后选取一定数量的测试样本对网络进行了测试,证明了系统的可行性.结果表明,该方法行之有效,在燃气轮机故障诊断领域中有很好的应用价值.

Research on expert system of gas turbine fault diagnosis based on fuzzy neural network

Considering the limitations of neural network and expert system in gas turbine fault diagnosis respectively, a new method of combining fuzzy neural network and expert system was proposed. The method could deal with the difficulty in knowledge accessing and decrypting fuzzy knowledge in traditional expert system. Typical training and testing examples were selected from developed simulator of three-shaft gas turbine in order to testify feasibility of the system. Result shows that this method is feasible. It has practical value in the field of gas turbine fault diagnosis.