Science-Guided AI for Development of New Biofuels and Bioenergy Production Technologies

M. Urgun-Demirtas, Y. Lin, P. Laible
Argonne National Laboratory,
United States

Keywords: process development, AI, modelling

Summary:

Meltem Urgun-Demirtas, YuPo Lin, Phil Laible * Argonne National Laboratory, 9700 S Cass Avenue, Lemont, IL, USA, demirtasmu@anl.gov Science-Guided AI for Development of New Biofuels and Bioenergy Production Technologies Multi-scale integrated unit operations modelling has been conducted for each unit operation in biofuels production to develop transfer functions to scale up the conversion process that can handle a variety of different feedstock characteristics. The initial data categorization and PCA execution help analysis of large data set, hence development of the frame work design and initial model elements setup; and then validation algorithm of decision making for AI. The system performance for a wide range of operating envelope will be studied, then performance range and relative frequency will be studied for conversion of highly variable feedstocks. Then, the correlation between feedstock properties and unit operation parameters will be further evaluated. Modeling based on the training data will be used to fit the model on the test data to validate the developed model at system level.