<![CDATA[Accurate load forecasting plays a critical role in ensuring the reliability and efficiency of electrical distribution systems. Increasing load variability, the integration of renewable energy, and changes in consumption behavior have intensified forecasting complexity. This study analyzes the effectiveness of time series methods for load forecasting in electrical distribution systems through a structured literature-based analytical approach. The study reviews and synthesizes empirical findings from peer-reviewed journal articles, conference proceedings, and patents published between 2002 and 2025. The methods analyzed include classical statistical models such as ARIMA, SARIMA, ARIMAX, and exponential smoothing, as well as hybrid and advanced approaches including LSTM, Prophet, fuzzy time series, wavelet-based models, and probabilistic forecasting frameworks. The results indicate that classical time series models remain effective for short-term forecasting with stable patterns, while hybrid and machine learning-based time series models provide superior performance under high volatility and complex load dynamics. Studies consistently report improvements in forecasting accuracy, measured using RMSE, MAE, and MAPE, when external variables and hierarchical structures are incorporated. The findings highlight the continued relevance of time series analysis as a foundational approach for load forecasting, while emphasizing the need for adaptive and hybrid models to address modern distribution system challenges. This study contributes a systematic synthesis that supports methodological selection for researchers and practitioners in electrical load forecasting.]]>
Copyrights © 2025
