Improved methods in statistical and first principles modeling for batch process control and monitoring
| dc.contributor.advisor | Edgar, Thomas F. | en |
| dc.creator | Zhang, Yang, 1980- | en |
| dc.date.accessioned | 2012-09-18T15:14:34Z | en |
| dc.date.available | 2012-09-18T15:14:34Z | en |
| dc.date.issued | 2008-08 | en |
| dc.description | text | en |
| dc.description.abstract | This dissertation presents several methods for improving statistical and first principles modeling capabilities, with an emphasis on nonlinear, unsteady state batch processes. Batch process online monitoring is chosen as a main research area here due to its importance from both theoretical and practical points of view. Theoretical background and recent developments of PCA/PLS-based online monitoring methodologies are reviewed, along with fault detection metrics, and algorithm variations for different applications. The available commercial softwares are also evaluated based on the corresponding application area. A detailed Multiway PCA based batch online monitoring procedure is used as the starting point for further improvements. The issue of dynamic batch profile synchronization is addressed. By converting synchronization into a dynamic optimization problem, Dynamic Time Warping (DTW) and Derivative DTW (DDTW) show the best performance by far. To deal with the singularity point and numerical derivative estimation problems of DTW and DDTW in the presence of noise, a robust DDTW algorithm is proposed by combining Savitzky-Golay filter and DDTW algorithm together. A comparative analysis of robust DDTW and available methods is performed on simulated and real chemical plant data. As traditional Multiway PCA-based (MPCA) methods consider batch monitoring in a static fashion (fail to consider time dependency between/within process variables with respect to time), an EWMA filtered Hybrid-wise unfolding MPCA (E-HMPCA) is proposed that considers batch dynamics in the model and reduce the number of Type I and II errors in online monitoring. Chemical and biochemical batch examples are used to compare the E-HMPCA algorithm with traditional methods. First principles modeling is known to be time consuming for development. In order to increase modeling efficiency, dynamic Design of Experiments (DOE) is introduced for Dynamic Algebraic Equation (DAE) system parameter estimation. A new criterion is proposed by combining PCA and parameter sensitivity analysis (P-optimal criterion). The new criterion under certain assumptions reduce to several available criteria and is suitable for designing experiments to improve estimation of specific parameter sets. Furthermore, the criterion systematically decomposes a complex system into small pieces according to PCA. Two engineering examples (one batch, one continuous) are used to illustrate the idea and algorithm. | en |
| dc.description.department | Chemical Engineering | en |
| dc.format.medium | electronic | en |
| dc.identifier.uri | http://hdl.handle.net/2152/17920 | en |
| dc.language.iso | eng | en |
| dc.rights | Copyright is held by the author. Presentation of this material on the Libraries' web site by University Libraries, The University of Texas at Austin was made possible under a limited license grant from the author who has retained all copyrights in the works. | en |
| dc.subject.lcsh | Electronic data processing--Batch processing--Statistical methods | en |
| dc.subject.lcsh | Electronic data processing--Batch processing--Mathematical models | en |
| dc.subject.lcsh | Chemical engineering--Mathematical models | en |
| dc.subject.lcsh | Algorithms | en |
| dc.title | Improved methods in statistical and first principles modeling for batch process control and monitoring | en |
| thesis.degree.department | Chemical Engineering | en |
| thesis.degree.discipline | Chemical Engineering | en |
| thesis.degree.grantor | The University of Texas at Austin | en |
| thesis.degree.level | Doctoral | en |
| thesis.degree.name | Doctor of Philosophy | en |