Copyright (c) 2025 REICIT
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This paper addresses load forecasting in electrical distribution networks using a methodology that is not limited to a single variable, but rather selects and analyzes multiple variables related to load prediction, using viable data from real-world network scenarios. Several future lines of research are proposed, such as the integration of the methodology into real-time systems to
monitor network variables and generate hourly forecasts, as well as the incorporation of intelligent systems that automatically respond to outages and allow queries on historical data and load factors. Furthermore, it is suggested that the models be improved using techniques such as MLP networks, genetic algorithms to optimize neural network parameters, and the joint application of neural models and wavelet transforms to filter noise in data series and improve prediction accuracy. These proposals seek to increase the effectiveness and applicability of the methodology in real-world operating environments.
Este obra está bajo una licencia de Creative Commons Reconocimiento-NoComercial-CompartirIgual 4.0 Internacional.
