基于概率PCA过程监控中遗失数据的重构(英文)

被引:6
作者
赵忠盖
刘飞
机构
[1] 江南大学自动化研究所
关键词
PPCA; 遗失数据; 数据重构;
D O I
10.16866/j.com.app.chem2006.12.008
中图分类号
TP277 [监视、报警、故障诊断系统];
学科分类号
0804 ; 080401 ; 080402 ;
摘要
在实际工业过程中,PCA常常被用于数据重构。但是相比于概率PCA(PPCA),PCA无论在建模上还是在统计监控指标上都存在一些缺陷。基于此,本文提出一种基于PPCA的遗失数据重构方法。通过使样本数据点与其在PPCA模型上的投影点之问的距离最小,该方法能够有效地进行数据重构。此外,还分析了使样本数据白化值最小的数据重构方法。在田纳西-伊斯曼过程中的应用验证了其有效性。
引用
收藏
页码:1205 / 1208
页数:4
相关论文
共 10 条
[1]  
Improved diagnosis of sensor faults using multivahate statistics. Dirk Lieftucht,et al. Proceeding of the 2004 American control conference . 2004
[2]  
Mixtures of principal component analyzers. Michael E.Tipping,Christopher M.Bishop. Neural Computation . 1999
[3]  
A unified geometric approach to process and sensor fault identification and reconstruction: the unidi-mensional fault case. Ricardo Dunia,and Qin S Joe. Computers and Chemistry . 1998
[4]  
Process Monitoring based on probabilistic PCA. Dongsoon Kim,In-Beum Lee. Chemometrics and Intelligent Laboratory Systems . 2003
[5]  
Missing data methods in PCA and PLS: Score calculations with incomplete observation. Philip R.,C Nelson, et al. Chemometrics and Intelligent Laboratory Systems . 2001
[6]  
Dealing with missing data: PartⅡ. B.Walczak,D. L. Massart. Chemometrics and Intelligent Laboratory Systems . 2001
[7]  
Fault detection and diagnosis in industrial systems. Chiang LH,Russell EL,Braatz RD. . 2001
[8]  
Monitoring batch processes using multi-way principal component analysis. Paul Nomikos,and MacGregor J F. American Institute of Chemical Engineen Journal . 1994
[9]  
Multivariate SPC chart for monitoring batch processes. Paul Nomikos,MacGregor J F. Technometrics . 1995
[10]  
Dealing with missing data: PartⅠ. Walczak B,and Massart D L. Chemometrics and Intelligent Laboratory Systems . 2001