MANAGING COMPLEXITY IN LARGE SCALE CONTROL SYSTEMS

This paper presents a data-based iterative tuning technique for large systems with multiple parameters. The method uses multivariate statistical regression methods to capture the dependencies between the system parameters and the quality measures determining the performance of the system. Nonlinear numerical optimization methods are applied for parameter tuning. Complementary tools to support the tuning procedure are suggested. Results from a simulated case study on a continuous pulp digester model are presented.