Electricity demand and renewable energy forecasting-single and hybrid predictive algorithms

Abstract Accurate prediction of electricity demand is essential for effective grid management in Bangladesh’s rapidly evolving energy sector. In this study, short-term month-ahead electricity consumption for the Bangladesh Power System is forecasted for September 2023 using ANN and ANFIS models, where the ANN model achieves superior performance with a MAPE of 1.01% compared to 1.36% for ANFIS. Additionally, a novel ARIMAX-ANN hybrid framework is employed to project hydroelectric generation from the 230 MW Karnaphuli hydropower plant over the period from 2021 to 2023. The ARIMAX algorithm demonstrates high precision, achieving an RMSE of 0.273 feet in predicting Kaptai reservoir levels. Seasonal variations are effectively captured, with peak hydro generation observed during July-September and reduced output from December to March. Furthermore,…