Una metodología para la predicción de la demanda de energía eléctrica a corto plazo (A Methodology for Short-Term Load Forecasting)

Jamer Jiménez Mares (jmares@uninorte.edu.co)1, Katheryn Donado Mercado (katherynd@uninorte.edu.co)1, Christian G. Quintero M. (christianq@uninorte.edu.co)1

1Universidad del Norte

This paper appears in: Revista IEEE América Latina

Publication Date: March 2017
Volume: 15,   Issue: 3 
ISSN: 1548-0992

Demand forecasting is important for electrical analysis development by utilities. It requires low error levels in order to reach reliability in electrical analysis. However, the demand for energy has dissimilar profiles variations depending on the type of day, weather conditions and geographical area. For this reason, it is necessary to group those curves showing similar behaviors and characterize them to establish which factors are significant for understanding. This paper proposes a novel methodology to identify those significant factors in the forecasting model for energy demand, and measure their effect on the Mean Absolute Percentage Error (MAPE) criterion and error performance. The experimental results show advantages of this methodology for zones with several behaviors in hourly power consumption.

Index Terms:
Short-time load forecasting, Neural Networks, Clustering   

Documents that cite this document
This function is not implemented yet.

[PDF Full-Text (762)]