Control of Continuous Mixed‐Solution Mixed‐Product Removal Crystallization Processes

Abstract

Continuous mixed-solution mixed-product removal (MSMPR) crystallization is considered. This process has been studied well, however, different aspects, in particular, process modeling, monitoring, and control remain challenging. An innovative approach for online measurement of the crystal size distribution is presented. Furthermore, unscented Kalman filtering is applied to overcome biased concentration measurement. Finally, a discrepancy-based control is applied to continuous MSMPR crystallization and its closed-loop performance is evaluated.

Multi-Rate Data Fusion for State and Parameter Estimation in (Bio-)Chemical Process Engineering

“…Here, particle filters [6,32] can represent a promising alternative. Furthermore, the outlined approach could be used to extend existing estimators for distributed parameter systems as found in the description of particle formation [33][34][35] and biotechnological processes [36][37][38], to reconstruct the systems states and parameters in presence of measurement delays or multi-rate measurements.…”

Simultaneous design and control of an industrial two-stage mixed suspension mixed product removal crystallizer

“…PI controllers have been used by Grosh and co-workers [32] to control moments of the CSD, for which an appropriate CSD sensor is required. Video microscopy for CSD measurement and discrepancy-based control was used by Geyyer and co-workers to successfully control the third moment of the CSD, in case of fouling of the concentration measurement probe [36]. Various model predictive control (MPC) realizations have been studied: neural network models have been used to predict the control states [37] in an industrial scale sugar crystallizer; the self organizing map (SOM) is also used to update parameters in the prediction model [38].…”

“…Multiple papers report the inadequacy or performance deterioration of traditional PID controllers due to the system nonlinearities [8,32,36,37]; Yang and co-authors [8] discourage the use of PID-type controllers due to the low capability of dealing with changing operating conditions. This literature shows that the control of CSD attributes can be successfully solved only by means of advanced process controllers (APC), able to incorporate kinetic crystallization information from measurements (PAT) [32,35,42] and/or a (first principle) process model [37,38,43,44].…”

“…Intensification by adding a second unit capable of automatically actuating on the CSD itself is another viable control strategy [45]. It also emerges that control schemes for continuous crystallization needs to be thoroughly designed [36], despite the fact that the control problem for continuous units is currently considered much easier than its batch counterpart [1]. We direct the reader to the book chapters [46,47] for a general overview of control strategies and challenges for batch and continuous crystallization.…”

Computing Crystal Size Distribution by Focused‐Beam Reflectance Measurement when Aspect Ratio Varies

“…Therefore, it is essential for CSD to be obtained online in manufacturing processes. Process analytical technologies (PATs) such as optical reflectance measurement [11][12][13], online image processing [14][15][16][17], focused-beam reflectance measurement (FBRM) [18][19][20][21], etc. are increasingly used for in-line measurement of CSD.…”