Automated Generation of Mechanistic Models for Chemical Process Digital Twins using Reinforcement Learning - Part II: Compartmentalization and Learning-Based Recalibration

Digital twins are virtual companions for the design, scale-up, and control of chemical processes. Equipping digital twins with mechanistic models of their mirrored unit operation expands their range of applicability compared to pure data-driven models. As constructing mechanistic models requires time, effort, and expert knowledge, automating their generation could make digital twins more accessible in the chemical industry. In this work, a workflow for automated generation of digital twins is extended to handle complex experimental systems comprised of interdependent, spatially distributed phenomena. The search for accurate models is performed by hierarchically connected reinforcement learning agents that operate on a basis of ontological system knowledge. The extended workflow is shown to reliably find accurate models of chemical systems, exemplified on a phase transfer catalysis reaction and a Taylor-Couette reactor. For the latter, its non-ideal flow patterns were predicted within a deviation of 5%, and automatically generated compartmentalizations were found to have comparable physical interpretations to bespoke models from literature. Additionally, the reinforcement learning agents were able to accurately recalibrate models up to two times faster when drawing upon pre-training under a different operation condition. By generalizing all parts of the automated modeling procedures, we enable the efficient (re-)use of knowledge previously confined to the human modeler. With this, the role of experts can be shifted from actively constructing each digital twin to being curators of knowledge for autonomous reinforcement learning agents.