While principal component analysis (PCA) has found wide application in process monitoring, slow and normal process changes often occur in real processes, which lead to false alarms for a fixed-model monitoring approach. In this paper, we propose two recursive PCA algorithms for adaptive process monitoring. The paper starts with an efficient approach to updating the correlation matrix recursively. The algorithms, using rank-one modification and Lanczos tridiagonalization, are then proposed and their computational complexity is compared. The number of principal components and the confidence limits for process monitoring are also determined recursively. A complete adaptive monitoring algorithm that addresses the issues of missing values and outlines is presented. Finally, the proposed algorithms are applied to a rapid thermal annealing process in semiconductor processing for adaptive monitoring.
|Number of pages||16|
|Journal||Journal of Process Control|
|Early online date||22 Aug 2000|
|Publication status||Published - Oct 2000|
|Event||14th IFAC World Congress - Beijing, China|
Duration: 5 Jul 1999 → 9 Jul 1999
Bibliographical noteFinancial support for this project from National Science Foundation under CTS-9814340, DuPont and Advanced Micro Devices is gratefully acknowledged.
- Adaptive process monitoring
- Lanczos tridiagonalization
- Rank-one modification
- Recursive principal component analysis