Pyomo.DOE: An open‐source package for model‐based design of experiments in Python

Abstract Predictive mathematical models are a cornerstone of science and engineering. Yet selecting, calibrating, and validating said science‐based models often remains an art in practice. Model‐based design of experiments (MBDoE) provides a systematic framework to maximize information gain from experiments while minimizing time and resource costs. But MBDoE remains limited to niche application areas, in part because practitioners must integrate expertise in statistics, computational optimization, and modeling. To help reduce this barrier, we introduce Pyomo.DOE , an open‐source package for MBDoE. Pyomo.DOE uses a nonlinear sensitivity analysis code k_aug to quickly approximate the Fisher information matrix and leverages a new stochastic programming abstraction. We demonstrate Pyomo.DOE with the first application of MBDoE to fixed‐bed breakthrough experiments, which highlights the power of Pyomo.DOE to quantify the value of experimental modifications a priori for large‐scale partial differential‐algebraic equation (PDAE) models. We also provide a mathematical primer on MBDoE targeted at general chemical engineers.