以信息融合和神经网络为理论基础,建立了风机融合神经网络模型和风机神经网络故障诊断模型。模型中,通过最近邻聚类学习算法对同类信息融合,应用人工神经网络对多源信息加以合成,并将融合结果作为输入层,进行神经网络故障诊断,实现了风机故障的智能诊断,并收到了良好效果。
This paper uses information fusion and neural net techniques as theoretical basis for building a fused neural-net model and a neural-net fault diagnosis model for a vent fan. In this model the congener information is fused by use of the most-closely clustering algorithm and the multi-source information is composed using the artificial neural-net technique. And then the fused results are used as an input source for neural-net fault diagnosis. In such a way, intelligent diagnosis is realized and the effect is good.