Recent Developments of the International Future Lab of Hydrogen Economy: Using Artificial Intelligence (AI) Based Machine Learning (ML) in r-SOC Development

Abstract

Current r-SOC research places special emphasis on the integrated approach of traditional numerical methods like CFD and accelerated AI methods. Using experimentally validated numerical models, the complex thermochemistry of auxiliary components containing syngas was evaluated. This method is used to generate data for the successful development and training of AI-based machine-learning models. Analyses shed light on the interactions between process variables in order to improve and prepare SOC-ready fuel, which is crucial for operation success. A recently developed ML model is utilised effectively to forecast and optimise reforming processes with various fuel constellations, including syngas compositions containing oxygen. Consequently, the results contribute to a greater understanding and qualitative benefits of preparing high-quality, pure syngas, improved fuel utilisation advancing sustainable research, and safe, consistent r-SOC operation. Consequently, early availability of valuable information is achieved. In addition, the strategy reduced prohibitive experiments, which contributed to the sustainable utilisation of resources.