Abstract:The digitalization of chemical research and industry generates large amounts of data, offering new opportunities for developing and parameterizing kinetic models. Leveraging this data requires machine learning techniques capable of autonomously extracting kinetics from reactor datasets. Recently, neural ordinary differential equations (neural ODEs) were coupled with reactor models to learn kinetics from ideal reaction systems, such as plug-flow reactors. However, real reactor set-ups commonly feature non-idealities such as heat- and mass transfer limitations, described by partial differential equations (PDEs) or differential-algebraic equations (DAEs). Discretizing such non-ideal PDE or DAE reactor models by finite volumes yields algebraic balance equations, which are solved by numerical schemes. In this work, we propose to learn kinetics from non-ideal reactor data using these balance equations and implicit neural networks, avoiding expensive backpropagation through the numerical solution by utilizing the implicit function theorem. The approach is demonstrated for the example of a mass transfer limited flat plate reactor for the preferential catalytic oxidation of CO. We show that Global Reaction Neural Networks, embedding thermodynamic and stoichiometric prior knowledge, coupled with a discretized two-phase CSTR cascade reactor model extract intrinsic kinetics from 50 integral reactor experiments. These kinetics generalize to new reactor geometries, where kinetics obtained by neural ODEs fail. Further, the approach is robust, recovering accurate kinetics even when training data is perturbed by 10% Gaussian noise. We expect that combining neural network-based kinetics with non-ideal reactor models will broaden the scope of kinetic model discovery and improve access to accurate kinetic models.

Iterative design of training data to control intricate enzymatic reaction networks

Active learning maps the emergent dynamics of enzymatic reaction networks.

Molecular system identification for enzyme directed evolution and design

Inferring Catalysis in Biological Systems

Forward design of a complex enzyme cascade reaction

Exploring Emergent Properties in Enzymatic Reaction Networks: Design and Control of Dynamic Functional Systems

Scalable Parameter Estimation for Genome-Scale Biochemical Reaction Networks

Iterative approach to computational enzyme design

An Efficient Approach to Bioconversion Kinetic Model Generation Based on Automated Microscale Experimentation Integrated with Model Driven Experimental Design

Exploring catalytic reaction networks with machine learning

Control Theory Concepts for Modeling Uncertainty in Enzyme Kinetics of Biochemical Networks

A versatile active learning workflow for optimization of genetic and metabolic networks

Benchmarking chemical neural ordinary differential equations to obtain reaction network-constrained kinetic models from spectroscopic data

Learning kinetics from non-ideal reactors by implicitly solved finite volumes and global reaction neural networks

The Automated Discovery of Kinetic Rate Models -- Methodological Frameworks

Modeling Enzyme Kinetics: Current Challenges and Future Perspectives for Biocatalysis

Quantitative Online Monitoring of an Immobilized Enzymatic Network by Ion Mobility-Mass Spectrometry

Biocatalysed synthesis planning using data-driven learning

Computational design of highly active de novo enzymes

Data-efficient modeling of catalytic reactions via enhanced sampling and on-the-fly learning of machine learning potentials

Heuristics-Guided Exploration of Reaction Mechanisms