Model-based Optimal Operation of Seeded Batch Crystallisation Processes

Dynamic optimization is applied for optimal control of a semi-industrial batch crystallisation process. The control strategy is based on a non-linear moment model. The dynamic model, consisting of differential and algebraic equations, is optimized using the simultaneous optimization approach in which all the state and input trajectories are parameterized. The resulting non-linear programming problem is then solved by a sequential quadratic programming algorithm. The optimal operation is realized by manipulation of the heat input to the crystallizer such that a maximum growth rate is maintained in the course of the process to avoid product quality degradation stemmed from irregular crystal growth. To be able to effectively track the maximum growth rate, the optimal heat input profile is computed on-line with feedback of process states estimated by an extended Luenbergertype observer; this enables the optimizer to reject the unmeasured process disturbances and also account for plant-model mismatch. Application of the on-line optimization strategy leads to substantial increase in the amount of crystals produced in a batch.