Energy Generation Forecasting in Biomass Power Plants: Integrating AI and Smart Grid Technologies

The energy transition concept, which is one of the important methods of mitigating climate change, refers to the rapid transition from fossil resources to renewable energy sources (RESs). Besides, new technologies such as electric vehicles and data centers, result in an increase in energy demand on the electric power side. The existence of more uncertainties and system components on the production and consumption side requires more attentive and efficient network management. As an alternative solution to the newly developing situation, the use of biomass energy, which is one of the RESs, in the electricity generation from landfill gas in areas with high population density is becoming a common technology. In this study, the generation forecasting of a biomass power plant, which works together with various power plants in meeting the energy needs of a region, is estimated. The algorithm considers the generation plans of other facilities, load demand and the variable market price. The data set of the system is used for production forecasting using various machine learning methods which are XGBoost, Decision Tree and Random Forest regression. The experimental results show that the XGBoost Regressor performs the best forecast, with a R² value of 0.9741.