Abstract
Batch operated crystallizers are widely used in the processing of fine chemical and pharmaceutical products. However, the tracking of various temperature setpoint trajectories appears to be a very difficult task if the crystallizer is cooled through vacuum, and advanced closed-loop control strategies have to be considered. This paper describes the application of the so called partial state model reference generalized predictive control approach (PSMR GPC) to the crystallization of organic compounds. After a brief introduction to the processes of crystallization from the solution, the main difficulties related to the use of vacuum for cooling industrial crystallizers are emphasized, followed by the weaknesses of PI control. The PSMR strategy is recalled and this algorithm is applied to calculate the input pressure leading to satisfactory tracking of temperature trajectories. The design of a non-linear adaptive supervisor is then described. In addition to the basic control law, this supervisor has been implemented to ensure improved safety and robustness. Supervisory PSMR GPC performances are reported. With respect to the main industrial constraints and requirements, the results are encouraging.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call