An intelligent Proportional controller of a Seeded Batch Crystallizer

Crystallization process is used in a very wide range of industrial applications. However, highly nonlinear comportment of such process and the difficulties of characterizing several phenomenological effects makes difficult to find suitable operational procedures for producing required products. In this article, we use a model-free control (MFC) for controlling the mean size of crystals produced by seeded batch cooling crystallization. The MFC method is distinguished notably in terms of the modeling strategy. Rather than developing a crystallization model within the classic population balance equation (PBE) together with the mass balance and the energy balance, as is usually done, we use a local approximation model. The model is continuously updated using the unique knowledge of the input-output behavior to handle both unknown dynamics and any possible disturbances. The MFC is only based on intelligent Proportional (iP) control. Simulation results prove the iP controller's efficiency even with noisy measurements, disturbances, and parameter variations.