摘要
当前检测造纸机械设备时,主要依托故障树模型进行故障分析,只能进行正向诊断推理,使得故障诊断结果F1值较低。因此,提出基于模糊贝叶斯网络的造纸机械设备故障诊断方法。将传感器设备分布式安装到造纸机械设备上,获取设备运行数据,并通过开/闭和闭/开滤波器消除噪声数据。考虑造纸机械设备工作体系,建立一个模糊故障树。从故障树的逻辑门入手,将其转化为包含多个故障事件节点的模糊贝叶斯网络,引入模糊集合论方法确定条件概率,构建基于模糊贝叶斯网络的故障推理模型,实现故障诊断反向推理。最后,对比条件概率与判断阈值,得出设备故障诊断结果。实验结果表明:所提方法诊断结果的F1值大于0.9,极大提升了故障诊断质量。
Fault analysis of paper machinery equipment mainly relies on the fault tree analysis model,which can only perform forward diagnostic reasoning and results in a low F1 value for fault diagnosis.Therefore,a paper machinery equipment fault diagnosis method based on fuzzy Bayesian network is proposed.Sensor devices are installed on the papermaking machinery equipment to collect operation data,and noise data is eliminated through open/close and close/open filters.Considering the working system of paper machine equipment,a fuzzy fault tree analysis is established.Starting from the logic gate of the fault tree analysis,it is transformed into a fuzzy Bayesian network containing multiple fault event nodes.The fuzzy set theory is introduced to determine conditional probability,and a fault reasoning model based on fuzzy Bayesian network is constructed to achieve reverse reasoning for fault diagnosis.Finally,the results of equipment fault diagnosis are obtained by comparing the conditional probability with the judgment threshold.Experimental results show that the F1 value of the proposed method's diagnostic results is greater than 0.9,significantly improving the quality of fault diagnosis.
作者
王丽丽
WANG Lili(Yancheng Polytechnic College,Yancheng 224000,China)
出处
《造纸科学与技术》
2023年第4期24-29,共6页
Paper Science & Technology
关键词
模糊贝叶斯网络
造纸企业
机械设备
故障诊断
故障树
条件概率
fuzzy Bayesian network
paper enterprises
mechanical equipment
fault diagnosis
fault tree analysis
conditional probability
