Abstract
In industrial process, the method based on principal component analysis (PCA) for data reconstruction is popular during dealing with missing data. In the method, the objective for data reconstruction is to make minimum square prediction error (SPE), an index measuring the relationship among process variables. However, PCA is a special case of factor analysis (FA), it has more limitations than FA, above all, SPE is a Euclidian distance, which is a suitable measurement for variables meeting uniform distribution rather than normal distribution, while process variables often satisfy the latter distribution. Due to the extensive sense of FA, the paper proposes a Mahalanobis distance as the index to measure the degree that the sample accords with the FA model, and then introduces FA into data reconstruction. The proposed index can more reflect the relationship among variables and the estimation of missing data with more precision can be achieved by making the new index minimum than by SPE.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Goulding, P.R., et al.: Fault Detection in Continuous Processes Using Multivariate Statistical Methods. International Journal of Systems Science 31(11), 1459–1471 (2000)
Ricardo, D., et al.: Use of Principal Component Analysis for Sensor Fault Identification. Computers and Chemical Engineering 20(suppl.), S713–S718 (1996)
Ricardo, D., et al.: Identification of Faulty Sensors Using Principal Component Analysis. AIChE Journal 10(10), 2797–2812 (1996)
Fuat, D., et al.: A Strategy for Detection and Isolation of Sensor Failures and Process upset. Chemometrics and Intelligent Laboratory Systems 20, 109–123 (2001)
Ricardo, D., Qin, S.J.: A Unified Geometric Approach to Process and Sensor Fault Identification and Reconstruction: the Unidimensional Fault Case. Computers and Chemical Engineering 22(7-8), 927–943 (1998)
Walczak, B., Massart, D.L.: Dealing with Missing Data: Part I. Chemometrics and Intelligent Laboratory Systems 58, 15–27 (2001)
Walczak, B., Massart, D.L.: Dealing with Missing Data: Part II. Chemometrics and Intelligent Laboratory Systems 58, 29–42 (2001)
Philip, R.C.N., et al.: Missing Data Methods in PCA and PLS: Score Calculations with Incomplete Observation. Chemometrics and Intelligent Laboratory Systems 35, 45–65 (2001)
Dirk, N., et al.: Improved Diagnosis of Sensor Faults Using Multivariate Statistics. In: Proceeding of the 2004 American control conference, Boston, Massachusetts, June 30-July 2, pp. 4403–4407 (2004)
Chiang, L.H., Russell, E.L., Braatz, R.D.: Fault Detection and Diagnosis in Industrial Systems. Springer, London (2001)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zhao, ZG., Liu, F. (2008). Data Reconstruction Based on Factor Analysis. In: Sun, F., Zhang, J., Tan, Y., Cao, J., Yu, W. (eds) Advances in Neural Networks - ISNN 2008. ISNN 2008. Lecture Notes in Computer Science, vol 5264. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-87734-9_56
Download citation
DOI: https://doi.org/10.1007/978-3-540-87734-9_56
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-87733-2
Online ISBN: 978-3-540-87734-9
eBook Packages: Computer ScienceComputer Science (R0)