Artificial Neural Network Based Short-Term Load Forecasting For the Ethiopian Electric and Power Corpora Tion-(Eepco)
No Thumbnail Available
Date
2004-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Load forecasting has become, in recent years, one of the major areas of research.
Most traditional forecasting and artificial intelligence researches have tried out this
task. Artificial neural networks (ANNs) have lately received much attention, and
successful experiments and practical tests have been reported. This work studies the
applicability of this kind of model procrastinating.The multi-layered feed-forward neural network, that are capable of representing nonlinear
fictional mappings between inputs and outputs was used to model the short term
load to recast for the Ethiopian electric and power corporation (EEPCO). The
network was trained with the error back-propagation method. Two models were
studied in this whole process. The first one is forecasting the load one hour ahead
and secondly the daily peak load forecast.The test results, based on historical demand, indicates that this methodology is
capable of providing accurate forecasts with 1.1 % and 1.3 % average absolute
forecast errors for the hourly and daily peak load forecasts respectively.
Description
Keywords
Information Science