摘要
提出了一种以扩展Park矢量方法为故障特征提取手段、利用BP网络的模式识别功能自动诊断异步电机转子断条故障的新方法。该方法消除了故障信号中基频成分对断条故障特征分量的“湮没”影响,同时实现了故障的自动识别,免去了人为介入。故障诊断实例表明:该方法具有良好的有效性和准确性。
A new method for automatically diagnosing broken bar fault of rotor of asynchronous motor was proposed in the paper, which taking extended Park's vector approach to pick up the fault characteristics and using BP network to recognize the mode of fault. When the method was applied, hiding effect of foundational frequency in fault characteristic was removed, and artificial interference was avoided. The results of practical example indicated that the method has good effectiveness and accuracy.
出处
《工矿自动化》
北大核心
2006年第6期8-11,共4页
Journal Of Mine Automation
基金
中国矿业大学科研基金资助项目(OC4499)
关键词
异步电机
转子断条故障
自动识别
BP神经网络
PARK矢量
asynchronous motor, broken-bar fault of rotor, automatic dagnosing, BP neural network, Park vector