OPTIMIZATION OF BATCH REACTOR OPERATION UNDER PARAMETRIC UNCERTAINTY - COMPUTATIONAL ASPECTS

The paper describes a method for optimizing batch reactors when the models at hand are characterized by parametric uncertainty. A discrete (or discretized) probability distribution of the uncertain parameters is assumed. This leads to a differential/algebraic optimization problem (DAOP) including several model descriptions, each corresponding to a grid point in parameter space. The DAOP is then transformed to an algebraic optimization problem (AOP) using a time parameterization based on the method of orthogonal collocation. This allows the user to (i) easily include additional algebraic path or endpoint constraints, and (ii) use a computationally-attractive simultaneous solution and optimization approach. Successive linear programming is used for solving the large and sparse AOP. The proposed solution strategy is illustrated on two optimization studies of a simulated batch reactor.