Data for “Neural network-based Grey-box Modeling and Economic Model Predictive Control of a Solvent-based Carbon Capture Pilot Plant

This resource helps you research and implement analysis and modeling of solvent‑based carbon capture using data from an experimental pilot plant. The dataset comes from the Danish Technological Institute pilot plant and includes: – Raw measurement logs – Neural‑network‑based grey‑box model parameter estimates for 11 different regularization values – Closed‑loop operational data from economic Model Predictive Control (eMPC) tests The main engineering challenge is to capture the complex, non‑linear thermodynamic and chemical behavior of carbon‑capture systems. By mixing physical equations with neural networks in a grey‑box setup, the model stays physically consistent while giving very accurate predictions. The closed‑loop test data also records real‑world performance, solver statistics, and the inputs

used by the economic MPC. M.Tech researchers can use this dataset to: – Validate models – Design controllers – Compare advanced control strategies Because the repository only contains existing experimental logs and parameter estimates, you cannot rebuild the pilot plant physically. The work should focus on simulation‑based validation, improving the model structure, and tuning controllers with the provided open‑source code and data.

Message: Hi FE, I need help with “Data for "Neural network-based Grey-box Modeling and Economic Model Predictive Control of a Solvent-based Carbon Capture Pilot Plant” in “Electrical Engineering”