Crystallization is used to produce vast quantities of materials. For several applications, continuous crystallization is often the best operation mode because it is able to reproduce better crystal size distributions than other operation modes. Nonlinear oscillations in continuous industrial crystallization processes are a well-known phenomenon leading to practical difficulties such that control actions are necessary. Nonlinear oscillations are a consequence of the highly nonlinear kinetics, different feedbacks between the variables and elementary processes taking place in crystallizers units, and the non-equilibrium thermodynamic operation. In this paper the control of a continuous crystallizer model that displays oscillatory behavior is addressed via two practical robust control approaches: (i) modeling error compensation, and (ii) integral high order sliding mode control. The controller designs are based on the reduced-order model representation of the population balance equations resulting after the application of the method of moments. Numerical simulations show good closed-loop performance and robustness properties.
- Crystallization; Continuous crystallization; population balances; nonlinear dynamics; robust model-based control.
Available at: http://works.bepress.com/hector_puebla/21/