摘要
转子是旋转机械的核心部分,对旋转机械故障的监测主要是对转子的运行状态检测;目前的转子运行监测与故障诊断大都采用依靠人工经验与定期巡检,不能依靠历史数据建立故障预测诊断模型;提出一种基于优化决策树的转子故障诊断方法,提取转子故障中丰富的频域信息,按照频率的组合构成决策树的分裂属性,采用决策树构造转子故障诊断模型;实验证明,该模型对转子的故障预测与诊断具有较高的准确率,具有很强的实用价值。
The rotor is the core part of rotating machinery, rotating machinery fault monitoring is the main rotor running state detection. The present rotor condition monitoring and fault diagnosis mostly rely on artificial experience and regular inspection, can't rely on the historical data to establish fault forecast and diagnosis model. Put forward a kind of decision tree based on Optimization of rotor fault diagnosis method of extracting rotor fault, rich in frequency domain information, in accordance with the frequency composition of decision trees split- ting attribute, using decision tree structure of rotor fault diagnosis model. Experiments show that, the model of the rotor fault prediction and diagnosis has high accuracy, has the very strong practical value.
出处
《计算机测量与控制》
北大核心
2013年第9期2375-2377,2381,共4页
Computer Measurement &Control
关键词
转子故障
决策树
频域信息
rotor fault
decision tree
frequency domain information