一种多组并联模糊神经网络用于信息融合诊断

Application of multi-information fusion for diagnosis in turbine generator set

研究了诊断对象多方位信息的综合利用,讨论了数据层融合的加权自适应最小平方估计、关联及特征层融合的模糊处理,并提出了一种新型的多组并联的模糊神经网络故障识别方法.在汽轮发电机组故障诊断中,解决了繁多种类的大量数据的快速处理,利用整体协调优势,提高了复杂设备并发故障诊断的精确度,并给出了应用实例.

The application of multi-information fusion can improve the precision and reliability of fault diagnosis in turbine generator set and other rotor machine. The comprehensive utilization of the multi-information fusion diagnosis object is studied. The method of fusion with weighted adaptive and least square and correlation at data level are researched and fusion of fuzzy set at characteristic level is discussed. A fault diagnosis method with multi parallel neural network are presented. In the fault diagnosis of turbine generator set, the problem of fast processing of various kinds of massive data was solved.