A techno-economic and ai-based optimization framework for hybrid energy systems supplying rural telecom base stations

Abstract

Abstract This paper introduces a strict AI-based framework of analysis of HRES in technical and economic dimensions to drive remote BTS.The proposed system delivers a total power output of 1.2 kW at − 48 V and 23 A, ensuring compatibility with standard telecom load requirements. A year’s worth of hourly simulation data is utilized to train and validate a range of forecasting algorithms, including linear regression, decision tree models, support vector machines (SVM), Gaussian process regression (GPR), kernel-based autoregressive moving average (KARMA) and feedforward neural networks (NN). EMS simulation results showed that hybrid solar-wind accounted for an average of 78.6% of the total daily load served, while fuel-based system usage was reduced by over 76% compared to conventional systems. The results confirm that intelligent forecasting and optimal dispatch strategies significantly improve system efficiency, reduce fossil fuel dependency and enhance the sustainability of HRES in decentralized telecom towers.