Risk mitigation in model-based experiment design: A continuous-effort approach to optimal campaigns

A key challenge in maximizing the effectiveness of model-based design experiments for calibrating nonlinear process models is inaccurate prediction information that afforded by each new experiment. We present a novel methodology to exploit prior probability distributions model parameter estimates bi-objective optimization formulation, where conditional-value-at-risk criterion considered alongside an average criterion. implement tractable numerical approach discretizes experimental space and leverages concept continuous-effort designs convex formulation. demonstrate tractability through three case studies, including dynamic experiments. In one case, Pareto frontier comprises campaigns significantly increase content worst-case scenarios. another same campaign proven be optimal irrespective risk attitude. An open-source implementation made available Python software Pydex.