摘要
研究基于机器学习算法,提出一种改进CNN网络的化工故障诊断方法。通过分析CNN网络结构特点与参数训练过程,采用PSO算法对CNN网络进行改进;然后,基于改进CNN网络,提出化工故障诊断方法;最后,通过以TE过程仿真软件,对本研究改进CNN算法在化工故障诊断中的应用进行验证。结果表明:可有效诊断化工故障,平均故障检出率达到91.23%,误报率为1.23%。相较于标准CNN算法、PCA算法、KPCA算法、MICA算法,改进CNN算法对化工故障的检出率更高,误报率更低,且故障检出速度更快。
A chemical fault diagnosis method based on machine learning algorithm and improved CNN network is proposed.by analyzing the structural characteristics and parameter setting method of CNN network.the PSO algorithm is used to improve the CNN network;Then based on the improved CNN network,a chemical fault diagnosis method is proposed.Finally,the application of the improved CNN algorithm in chemical fault diagnosis is verified by TE process simulation software.The results show that the method can effectively diagnose the chemical fault,the average fault detection rate is 91.23%,and the false alarm rate is 1.23%.Compared with the standard CNN algorithm,PCA algorithm,KPCA algorithm and MICA algorithm,the improved CNN algorithm has higher detection rate,lower false alarm rate and faster fault detection speed.
作者
许洪光
李凤英
郭茜
XU Hongguang;LI Fengying;GUO Qian(Hebei Oriental University,Langfang 065000,Hebei China)
出处
《粘接》
CAS
2022年第5期85-89,94,共6页
Adhesion
