基于信息融合技术的电机故障诊断

Motor fault diagnosis based on information fusion technology

为了能够从多方面反映电机系统状态,实现对电机故障模式的自动识别与准确诊断,将数据融合技术与神经网络相结合,建立电机故障诊断系统。在数据融合级上,将故障特征量进行分类处理,然后采用多层神经网络进行故障特征级融合与电机故障的局部诊断,获得彼此独立的证据,再运用D-S(DempserShafer)证据理论融合算法对各证据进行融合,最终实现对电机故障的准确诊断。诊断测试试验证明,该诊断系统提高了电机故障诊断的精度,并能满足诊断的实时性要求。

The mine motor fusion diagnosis system was set up for reflecting the mine motor system state in multi-aspect, realizing automatically the identification of motor fault modes, and diagnosing accurately the faults by using neural network and evidence theory. After fault characteristic data was classified and processed on the data fusion level, a multi-level neural network was used to carry on the fusion on the characteristic level and the local fault diagnosis of mine motors, so that independent evidences could be acquired. Then D-S evidence theory fusion algorithm was used to fuse all of the evidences to finally fulfill an accurate fault diagnosis for mine motors. The diagnosis tests prove that the system can improve the diagnostic precision and satisfy the requirement for real-time diagnosis.