کنفرانس بین المللی فناوری و مدیریت انرژی

صفحه اصلی / نهمین کنفرانس بین المللی فناوری و مدیریت انرژی

GlorEST: A Demand Power Forecasting Automated Machine Learning Tree-based Pipeline Optimization Predictive Precise Model

نویسندگان :

Ashkan Safari¹ Amir A.Ghavifekr² Mahmood Seyyedzadeh³ Alireza Tajdid⁴

1- University of Tabriz 2- University of Tabriz 3- University of Tabriz 4- University of Tabriz

کلمات کلیدی :

Demand Response،Predictive Modeling،Smart Power Systems،Energy Forecasting،Energy Efficiency;

چکیده :

Predictive AI-driven models play a considerable role in anticipating and managing electricity demand within smart power systems. By employing sophisticated algorithms to analyze historical consumption patterns, real-time data, and external factors, these models facilitate accurate forecasting crucial for optimizing energy distribution and enhancing grid reliability. This paper introduces GlorEST, an advanced modeling framework specifically tailored for the German electricity market. Utilizing a comprehensive dataset, GlorEST incorporates key features such as conventional and solar power supply, demand peaks, renewable energy contributions, trade dynamics, storage, and efficiency metrics. The simulation section dynamically captures the intricate interplay of these factors, offering a holistic view of the market dynamics. GlorEST's accuracy is validated through key performance indicators (KPIs) with a low Mean Absolute Error (MAE) of 0.0489, a high R-squared (R2) of 0.9941, and precise predictions reflected in the low Mean Absolute Percentage Error (MAPE) of 0.8691. Comparative analysis against LightGBM, XGBoost, and KNN underscores GlorEST's superior performance, positioning it as a reliable and versatile tool for decision-making and scenario analysis in the dynamic landscape of the German electricity market.