The substantial impact of buildings on CO2 emissions and climate change highlights the urgent need to adopt sustainable measures. Microalgae photobioreactors have shown potential as a renewable feedstock for biomass generation and CO2 absorption. Conventional biomass prediction methods are usually laborious and time-intensive, and sacrifice biomass for measurement. This research addresses these challenges by developing a novel approach using Feed-forward Neural Networks and Convolutional Neural Networks for accurate biomass prediction. Feed-forward Neural Networks demonstrated a positive linear relationship between predicted and valid biomass values of 0.99 on average. Convolutional Neural Networks improved performance over time with an accuracy of up to 97%. This research advanced biomass prediction methodologies and supported optimizing photobioreactors performance for enhanced energy efficiency and biomass generation.
