A review on model-based design of experiments for parameter precision – Open challenges, trends and future perspectives

In the last decades, the systematic use of mathematical models has become pervasive within the chemical process engineering context. Experiment design plays a critical role for the rapid calibration and validation of mathematical models – whether mechanistic, data-driven or semiempirical models – with the ultimate goal of driving process development and optimisation. Since data is often limited, and running additional experiments to gather more information for model calibration might be expensive or impractical, model-based design of experiments (MBDoE) techniques have become increasingly relevant to support the model calibration task on a broad range of process engineering applications. In this review, we aim to provide a comprehensive overview of the advances in the field that have occurred since the previous work presented by Franceschini and Macchietto (2008), Chemical Engineering Science, 63, 4846–4872. We first provide the theoretical foundations behind the standard MBDoE problem formulation and highlight recent advances; we then thoroughly analyse limitations, open challenges, and current trends by discussing about 250 contributions in the field. Finally, we highlight future research directions that may enlarge and further enhance the robustness and reliability of MBDoE implementation in real industrial scenarios.