基于双层机器学习的动态精馏过程故障检测与分离

被引：10

作者：

毛海涛

田文德

梁慧婷

机构：

[1] 青岛科技大学化工学院

来源：

过程工程学报 | 2017年 / 17卷 / 02期

关键词：

机器学习; 动态精馏过程; 故障检测与分离; 网络结构解析;

D O I：

暂无

中图分类号：

TP181 [自动推理、机器学习]; TQ028.31 [];

学科分类号：

140502 [人工智能];

摘要：

提出了基于双层机器学习的动态精馏过程故障检测和分离的方法,检测的阈值为正常工况训练的网络输出值与样本的残差.通过对比网络预测值和实测值的偏差检测故障,检测到故障时,启动另一网络对动态过程自适应拟合异常工况数据.网络的预测值与实测值的偏差小于阈值时,拟合成功.通过对两个网络进行结构解析找到造成输出变量异常波动的输入变量.将该方法运用到脱丙烷精馏塔中,检测出过程中的故障,并分离出与故障源相关的变量,表明该方法准确、有效.

引用

页码：351 / 356

页数：6

共 11 条

[1]

Fault detection in dynamic systems using the Kullback–Leibler divergence.[J].Lei Xie;Jiusun Zeng;Uwe Kruger;Xun Wang;Jaap Geluk.Control Engineering Practice.2015,

[2]

Optimizations of the Gravitationally Organized Related Mapping ANN through Genetic Algorithms.[J].Chris Gorman;Iren Valova.Procedia Computer Science.2014,

[3]

Fault detection and isolation of a dual spool gas turbine engine using dynamic neural networks and multiple model approach.[J].Z.N. Sadough Vanini;K. Khorasani;N. Meskin.Information Sciences.2014,

[4]

Fault detection and diagnosis for distillation column using two-tier model [J].

Tian, Wen-De ;

Sun, Su-Li ;

Guo, Qing-Jie .

CANADIAN JOURNAL OF CHEMICAL ENGINEERING, 2013, 91 (10) :1671-1685

[5]

Model selection and fault detection approach based on Bayes decision theory: Application to changes detection problem in a distillation column [J].

Chetouani, Yahya .

PROCESS SAFETY AND ENVIRONMENTAL PROTECTION, 2014, 92 (03) :215-223

[6]

Fault detection and diagnosis for non-Gaussian stochastic distribution systems with time delays via RBF neural networks [J].

Qu Yi ;

Li Zhan-ming ;

Li Er-chao .

ISA TRANSACTIONS, 2012, 51 (06) :786-791

[7]

Fault diagnosis and classification based on wavelet transform and neural network.[J].Kamal Hadad;Meisam Pourahmadi;Hosein Majidi-Maraghi.Progress in Nuclear Energy.2010, 1

[8]

An integrated method of independent component analysis and support vector machines for industry distillation process monitoring [J].

Bo, Cuimei ;

Qiao, Xu ;

Zhang, Guangming ;

Bai, Yangjin ;

Zhang, Shi .

JOURNAL OF PROCESS CONTROL, 2010, 20 (10) :1133-1140

[9]

An expert system design for a crude oil distillation column with the neural networks model and the process optimization using genetic algorithm framework.[J].S. Motlaghi;F. Jalali;M. Nili Ahmadabadi.Expert Systems With Applications.2007, 4

[10]

Selecting appropriate control variables for a heat-integrated distillation system with prefractionator [J].

Engelien, HK ;

Skogestad, S .

COMPUTERS & CHEMICAL ENGINEERING, 2004, 28 (05) :683-691

← 1 2 →