Optimization of second-generation ethanol production employing machine learning techniques

• Grant number: 24/13885-5
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2024
• Status: Discontinued
• Field of knowledge: Engineering - Chemical Engineering - Chemical Process Industries
• Principal Investigator: Antonio José Gonçalves da Cruz
• Grantee: Augusto Demetrio Canutilho
• Host Institution: Centro de Ciências Exatas e de Tecnologia (CCET). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil
• Associated scholarship(s): 25/01916-6 - Artificial neural networks as a surrogate model for simulating biomass enzymatic hydrolysis step, BE.EP.IC

Abstract

Biofuels have significant potential to diversify Brazil's energy matrix, contributing to the reduction of pollution caused by the combustion of fossil fuels. Sugarcane, the main raw material for the production of first-generation ethanol (E1G), generates by-products such as bagasse and straw, which can be used in the production of second-generation ethanol (E2G). Utilizing these residues allows for an increase in ethanol production without the need to expand cultivation areas, thus reducing pressure on arable land and greenhouse gas emissions. However, the production of E2G still faces challenges in achieving economic viability, requiring additional stages of pre-treatment and hydrolysis of biomass to obtain fermentable sugars. Therefore, this undergraduate research project proposes the development of software using artificial intelligence and machine learning techniques to identify optimal operational conditions for pre-treatment and enzymatic hydrolysis stages, aiming to maximize E2G production from sugarcane straw and bagasse. The computational tool, implemented in Python, will feature a graphical interface for visualization and interaction with production parameters (such as solid concentration, enzymes, and reaction time), and will be designed to be reusable and expandable, allowing for the incorporation of new models and experimental data to enhance its predictive capability.