Multiple Objective Optimisation of Cyclic Chemical Systems with Distributed Parameters

The optimisation of innovative cyclic chemical processes may be challenging because these problems often involve PDE models, temporal periodicity constraints, and multiple objectives. Therefore, this paper presents two deterministic approaches which allow a fast and efficient generation of the Pareto set for these multiple objective optimisation problems. Hereto, the Normal Boundary Intersection (NBI) and the Normalised Normal Constraint (NNC) method are incorporated into a deterministic multiple shooting approach. Both approaches are illustrated for a reverse flow reactor case study involving conflicting conversion and energy objectives, and their performance is compared to a classic weighted sum approach.