神经网络在汽轮机转子故障诊断中的应用研究  

Application Research of Neural Network in Fault Diagnosis of Turbine Rotor

作　　者：高树奎 华兴鲁 孟帅 Gao Shukui;Hua Xinglu;Meng Shuai(Handan Cogeneration Power Plant, Handan 056004, China;Matou Power Plant, Handan 056044, China)

机构地区：[1]国电电力邯郸热电厂,河北邯郸056004 [2]马头发电厂,河北邯郸056044

出　　处：《河北电力技术》2018年第1期29-32,共4页Hebei Electric Power

摘　　要：针对汽轮机转子在运行过程中出现故障的几率较高且危害较大,传统诊断方法费时费力的问题,建立BP神经网络及RBF神经网络转子故障诊断模型,通过实例对转子故障进行诊断,结果表明BP神经网络和RBF神经网络诊断与转子揭缸检查一致,但BP神经网络比RBF神经网络对转子的故障诊断精度更高,可为转子检修及安装设计提供了参考。Aiming at the high failure rate and harmfulness of turbine rotor during operation,the traditional diagnostic methods are time-consuming and labor-intensive,and BP neural network and RBF neural network rotor fault diagnosis model are established. The rotor fault diagnosis is carried out through examples. The diagnostic results of both The diagnosis of BP neural network and RBF neural network is consistent with the inspection of rotor uncovering cylinder. However,BP neural network is more accurate than RBF neural network in fault diagnosis of rotors,which provides a reference for rotor overhaul and installation design.

关 键 词：汽轮机转子 BP神经网络 RBF神经网络 故障诊断 

分 类 号：TK269[动力工程及工程热物理—动力机械及工程]