Robust Optimal Temperature Operation for Seeded Batch Crystallization Processes

Robust optimal temperature trajectories for seeded batch crystallization processes under size-dependent crystallization kinetics are proposed, which minimize the amount of fine crystals. First, a deterministic optimal temperature-swing trajectory is calculated by Particle Swarm Optimization (PSO). Furthermore, robust trajectories are proposed by considering parametric perturbations in the process model and initial conditions. Two robust versions of PSO, named AMPSO and WMPSO, are introduced to cope with the uncertain optimization problem. The result of Monte Carlo simulation demonstrates that the optimal trajectories are capable of producing fine-minimized crystals with high robustness against initial operating condition changes.