NONLINEAR MODEL PREDICTIVE CONTROL STRATEGIES APPLIED TO A FED-BATCH SUGAR CRYSTALLIZER
Systematic methods and tools for managing the complexity
Process Control (T4-8P)
Keywords: nonlinear model predictive control, fed-batch crystallization process
During the last decade the model based predictive control (MPC) became an attractive control strategy implemented in a variety of process industries. However, it can be considered as industrial alternative only for continuous and predominantly linear processes (Qin and Badgewell, 2003). The application of MPC for batch nonlinear cases is still far from being an industrial reality and represents an interesting theoretical and practical control challenge (Balasubramhanya and Doyle, 2000). The batch or fed-batch mode is a typical production scheme for a large group of pharmaceutical, biotechnological, food and chemical processes. It is related with the formulation of a control problem in terms of economic or performance objective at the end of the process (Nagy and Braatz, 2003). For example, the crystallisation quality is evaluated by the particle size distribution (PSD) at the end of the process which is quantified by two parameters - the final average (in mass) particle size (MA) and the final coefficient of particle variation (CV). The main challenge of the batch production is the large batch to batch variation of the final PSD. This lack of process repeatability is caused mainly by improper control policy and results in product recycling and loss increase. MPC, being one of the approaches that inherently can cope with process constraints, nonlinearities, and different objectives derived from economical or environmental considerations, has the potential to overcome the problem of the lack of repeatability and drive the process to its optimal state of profit maximization and cost minimization.
The present work is focused on a comparative analysis between two Nonlinear MPC (NMPC) schemes implemented to a batch white sugar crystallizer – i) NMPC that does not exploit the batch nature of the process (termed as classical NMPC) and ii) the batch NMPC that takes into account the end-point control objectives, i.e. the prediction horizon is equal to the batch final time. Three control scenarios were studied. In the first scenario the nominal case without disturbance and noise is considered. In the second and the third scenarios the effect of disturbance in the vacuum pressure and the noise in the vacuum pressure were studied.
The results demonstrated that in general the batch NMPC outperforms the classical NMPC but to the expense of heavy computational burden due to the high prediction horizon.
References
Balasubramhanya, L.S., Doyle III, F.J. Nonlinear model-based control of a batch reactive distillation column. Journal of Process Control, Vol. 10 (2000) 209-218
Nagy, Z.K. and R.D. Braatz (2003). Robust nonlinear model predictive control of batch processes. AIChE J., 49, 1776-1786.
Qin S. J., and T. Badgewell (2003). A Survey of Industrial Model Predictive Control Technology. Control Engineering Practice, 11, 733-764.
See the full pdf manuscript of the abstract.
Presented Tuesday 18, 13:30 to 15:00, in session Process Control (T4-8P).
