Optimal control of a freeze Dryer—II: Dynamic analysis

Abstract

The unsteady state model and Lagrangian equations of the freeze-drying process presented in Part I[4], have been solved for non-fat reconstituted milk and turkey, using the control strategies presented in [4]; the results were then used to evaluate the coefficients representing the necessary conditions of optimality. The results for non-fat reconstituted milk confirm the results of the quasi-steady state analysis [4], except for a time segment very close to t = 0. Simulations with turkey breast showed similar results at low pressures, but it was found that at higher pressures the assumed control policies which were based on a quasi-steady state analysis[4], were not optimal The non-steady state results indicate that for turkey, intermediate rather than maximum, pressure control policies were optimal for cases (a) and (b) in[4]. However, in practice these optimal pressure policies might produce only minor improvements in drying time when compared with a near optimal policy. It was also found that changes in the interface constraint affected drying times only marginally, but that large heaters operating under optimal control had a pronounced effect on the drying time.