<![CDATA[Electric load forecasting is a critical step in ensuring the reliability of power systems amid rising energy demand driven by digitalization, industrialization, and urbanization. This article presents a Systematic Literature Review (SLR) on the application of Artificial Neural Networks (ANN) with backpropagation algorithms for load prediction based on historical data, employing the PRISMA framework for study screening and selection. The review analyzes nine relevant national journals to identify trends in accuracy, network configurations, and model effectiveness. Findings indicate that ANN with backpropagation can achieve low prediction error rates, such as a Mean Absolute Percentage Error (MAPE) of 0.05% in industrial sectors and up to 99.88% accuracy in specific cases. ANN also demonstrates strong capability in capturing dynamic changes in energy consumption, making it a reliable method for supporting operational planning and efficient electricity distribution. Despite promising performance, several aspects remain underexplored, including more complex ANN architectures, hyperparameter tuning techniques, limited cross-regional validation, and insufficient comparative analysis with alternative methods such as ensemble learning or deep learning-based algorithms. This review offers comprehensive insights into the integration of artificial intelligence in power systems and lays the groundwork for developing more adaptive, precise, and broadly generalizable load forecasting strategies in the future.]]>
