摘要
汽轮发电机组作为电力行业的重要设备,其结构日益复杂,参数高、容量大,一旦发生故障将会产生更严重的影响。由于SOM神经网络更接近于人脑的认知规律,在对其算法进行研究的基础上,将其应用于汽轮发电机组振动故障诊断。通过算例仿真测试,可知该方法在汽轮发电机组振动故障诊断中具有较高的正确率,可将其集成在汽轮发电机组监测和故障诊断系统中,从而对提高机组的设备可靠性发挥更加积极的作用。
In this paper,SOM neural network,which is more close to the brain＇s cognitive law,is applied to fault diagnosis of turbine-generator unit based on the study of its algorithm.The high accuracy of method is proved by the simulation test of steam turbine unit vibration fault diagnosis.SOM neural network can be integrated in the turbo-generator set monitoring and fault diagnosis system,which plays a more active role in improving the reliability of the unit.
出处
《工业控制计算机》
2016年第1期137-139,共3页
Industrial Control Computer
关键词
SOM神经网络
汽轮发电机组
故障诊断
SOM Neural Network
steam turbine generator unit
fault diagnosis
