【摘要】 针对动密封中唇形密封的综合故障问题,应用小波神经网络技术加混合式数据融合方法,将数据级、特征级和决策级故障诊断数据融合在一起,对动密封中唇形密封的缺陷进行智能诊断,描述了小波神经网络的建模过程,探讨了通过多源互补信息减少故障诊断系统不确定性的优化方法。结果表明,采用混合式融合结构可以通过多源互补以及冗余信息来提高诊断系统的鲁棒性;采用小波神经网络信息融合的诊断方法,可有效诊断动密封中唇形密封存在的综合故障问题。更多还原

【Abstract】 The synthetical symptoms of labiate airproof in moving airproof was concluded.Using wavelet and mixed data merge method,which is integrated with data,characteristic,decision grate and nerve network,the intelligence diagnosis on the defect of labiate airproof was made.A model of wavelet neural network was constructed.In order to reduce no confirm of defect analysis,the excellent diagnosis way was studied with the information of many sources fill and redundant.The result shows that using mixed data merge may raise robustness with the help of many sources fill and redundant,and using wavelet and mixed data merge can effective diagnose the synthetical symptoms of labiate airproof in moving airproof.更多还原