摘要
将数据融合方法引入到异步电机的故障诊断中,通过不同传感器综合采集异步电机运行时的各个状态参数,运用并行BP子神经网络对异步电机进行局部诊断,再用D-S证据理论对局部诊断的结果进行全局融合,实现对异步电机的准确诊断。实验表明,诊断结果的可信度显著提高,不确定性明显减少,证明了该方法是有效的。
A fault diagnosis of asynchronous machine based on data fusion is presented, Several parameters are gotten by the multi-sensor system of comprehensive monitoring on the asynchronous machine, Then, two shunt-wound BP sub-networks are used to carry on local fault diagnosis, Meanwhile, the D-S evidence theory were used fuse those local diagnosis results, The simulation results show that there liability diagnostic improve while its uncertainty decrease prominently, The validity of this method has been proved significantly.
出处
《电气传动自动化》
2008年第2期34-36,共3页
Electric Drive Automation
