The objective of this project is to develop optimal strategies for controlling and monitoring multivariate continuous processes, following and approach which integrates Automatic Process Control (APC) and Statistical Process Control (SPC). This strategy minimises output variability and maximises productivity in the face of dynamically related observations, by making regular adjustments to one or more input and process parameters. The controller designs are developed using multi-input/multi-output transfer functions, based on the model-based predictive controller (MBPC) approach. The detection and elimination of assignable causes of variation can be achieved by monitoring process outputs, which are based on statistical tools for detecting outliers in multivariable ARIMAX models. The controller and monitoring procedure designs developed will be applied to actual continuous industrial processes from our collaboration with several companies.
KEYWORDS: Statistical Process Control (SPC), Automatic Process Control (APC), ARIMAX Models, Continuous Quality Improvement, Predictive control, Optimisation.