Email: [email protected]
PhD Scholar UCoBs Subproject 3-Renewable Energy (Biomass Option)
Research Title: Predicting the Quality of Biomass Briquettes from Agricultural and Municipal Solid Organic Waste Using Machine Learning Models.
About the Research
The research is meant to Develop and train machine learning models to predict key quality parameters of biomass briquettes (e.g., heating value, density, durability) based on the composition and properties of agricultural and Municipal Solid Organic Waste feedstock. Most influential factors affecting briquette quality will be identified using feature selection techniques. The research also involves optimization of briquette production parameters (e.g., binder type, pressure, moisture content) to improve quality based on model predictions.
Methodology:
The research involves:
- Collection of data on briquette properties (density, durability, calorific value) and corresponding waste composition and processing parameters.
- Production of briquettes with varying compositions and production parameters.
- Measurement of key briquette quality parameters (heating value, density, durability, compressive strength.).
- Development and training of machine learning models (e.g., Random forest, Extra Trees and ANN) using measured briquette quality as the target variable and feedstock properties as input features.
- Validation and optimization of models using statistical and performance metrics.
- Determination of the best stoves that can work efficiently with the briquettes produced during the study and hence foster the adoption of biomass briquettes as an energy source in Western UgandaResearch Supervisors:
Name Institution 1. Dr. Simon Kawuma Faculty of Computing and Informatics: MUST 2. Dr. Okot David Kilama Faculty of Science: MUST 3. Assoc. Prof Julien Blondeau Vrije Universiteit Brussel(VUB), Brussels, Belgium 4. Assoc. Prof Johan De Greef Faculty of Engineering Technology – KU Leuven, Belgium
