Research

Economic Dispatch in Power Systems with Intermittent Supply

Project Details

Abstract

This paper introduces a novel real-time optimization algorithm that maximizes market surplus by leveraging forecasts of grid intermittency. The algorithm integrates machine learning models and optimization techniques to predict blackout schedules and optimize generation dispatch and storage scheduling accordingly. Three machine learning models—Naïve, Long Short-Term Memory (LSTM), and Dynamic LSTM (DLSTM)—are evaluated for their predictive accuracy and their effectiveness in managing dynamic energy distribution. The primary innovation of this research lies in its real-time adjustment mechanism, which continuously aligns predictions with actual grid conditions, thus enhancing both the economic efficiency and reliability of the power system. The efficacy of the proposed algorithm is validated through a case study in Kabrikha, southern Lebanon, demonstrating its potential to significantly reduce operational costs and improve reliability in regions experiencing frequent power disruptions. Future enhancements will explore the integration of real-time pricing and demand response strategies to further optimize energy management and economic outcomes.