Abstract
Accurate energy forecasting is critical for the stability, efficiency, and cost-effectiveness of modern power grids, particularly as renewable energy sources like solar and wind become more prominent. The variability of these sources presents challenges for traditional forecasting models, which struggle with non-linearity, external dependencies, and evolving grid conditions. These limitations lead to operational inefficiencies, grid instability, and increased financial risks. This study investigates the effectiveness of traditional statistical models, advanced machine learning techniques, and hybrid forecasting approaches to enhance prediction accuracy and grid performance. Using real-world datasets from Kaggle, including historical energy generation, consumption, pricing, and weather variables, this research follows the CRISP-DM framework for data preprocessing, analysis, model selection, and evaluation. Forecasting models such as ARIMAX, Theta, LSTM, TCN, RF-LightGBM, and LSTM-TCN were tested across multiple time horizons. The results demonstrate that hybrid models integrating machine learning and statistical methods significantly outperform conventional approaches, offering greater precision and adaptability in forecasting volatile energy demand and generation patterns. Beyond model accuracy, the study examines the economic and operational impacts of forecasting inaccuracies, highlighting how poor predictions lead to higher costs, inefficient resource allocation, and increased reliance on backup power. Improved forecasting reduces these inefficiencies, enhances grid reliability, and optimizes energy distribution. Findings emphasize the importance of integrating real-time external variables, such as weather and market dynamics, to improve forecast reliability. This research contributes to advancing AI-driven energy forecasting by demonstrating its practical benefits in grid optimization. Future research should focus on improving model interpretability through explainable AI, optimizing computational efficiency for real-time deployment, and incorporating additional external factors such as policy changes and energy market fluctuations. Expanding dataset diversity and adapting forecasting models for decentralized energy systems and microgrids are also recommended. The insights gained provide valuable guidance for energy policymakers, grid operators, and researchers in developing more accurate, resilient, and cost-effective forecasting solutions.
Library of Congress Subject Headings
Renewable energy sources--Forecasting; Electric power systems--Forecasting; Electric power distribution--Reliability
Publication Date
5-2025
Document Type
Thesis
Student Type
Graduate
Degree Name
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research
Advisor
Sanjay Modak
Advisor/Committee Member
Khalil Al Hussaeni
Recommended Citation
Alboom, Rashed Khalid, "Energy Forecasting Inaccuracies and Their Direct Impact on Grid Performance" (2025). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/12057
Campus
RIT Dubai
Plan Codes
PROFST-MS
