Multi-objective optimisation of the algae Arthrospira platensis under uncertainty: a Pareto ellipsoids approach

Multi-objective optimisation allows a decision maker to evaluate trade-offs between objectives by supplying a set of optimal solutions, for example to find the balance between process performance and sustainability in process design. Accurate prediction of these objectives and, more importantly, safe operating conditions are needed. Thus, we need to account for uncertainty within the model. The Pareto ellipsoids approach ensures efficient use of computation while visualising the effect of parametric uncertainty. The novel Pareto ellipsoids-based algorithm is applied to the algae Arthrospira platensis and determines optimal light conditions for both biomass and product yield. Uncertain parameters are selected to investigate propagation of uncertainty from both the light intensity model and the Droop model. Finally, the obtained Pareto ellipsoids’ accuracy is verified using a Monte Carlo approach.