Physics-constrained ML surrogates for a CO2-to-methanol reactor: cleaning noisy plant logs (simulated with a kinetic model), exact mass balances and equilibrium limits that keep errors low outside the training range, and an optimiser for operating points.
python neural-network scikit-learn scipy ensemble-learning chemical-kinetics hydrogen data-cleaning catalytic-reactors energy-systems process-optimization surrogate-model power-to-x physics-informed-machine-learning methanol co2-utilization
-
Updated
Oct 15, 2026 - Python