Services on Demand
Journal
Article
Indicators
- Cited by SciELO
Related links
- Similars in SciELO
Share
Revista Científica de la UCSA
On-line version ISSN 2409-8752
Abstract
BARBOZA, OA. Hourly Load Forecast Automation in the National Interconnected System. Rev. ciente. UCSA [online]. 2014, vol.1, n.1, pp.4-14. ISSN 2409-8752.
The electric and energetic operation of the National Interconnected System should be planned to ensure the power supply with technical proficiency and low costs. This operational planning should consider technical constraints, such as equipment maintenance and generators unavailability, contractual issues and so on, as well as economic factors such as weather and holiday dates that affect the use of energy by various consumer groups. In this sense, the overall power demand of the country affects the hiring of hydroelectric power, which is the main component of the operating cost of ANDE. In this context, the efficient operation of the transmission system from the technical and economic terms, requires an accurate estimate of the demanded global power in the network, for each hour of each day (demand profile) and consideration of certain restrictions, making it necessary to use reliable methodologies for estimating such demand and dispatch the appropriate power from each power generation plant. This paper analyzes the various methodologies used for Short Term Load Forecast. They are applied a statistical approach (ARIMA) and artificial intelligence approach (Artificial Neural Network: ANN) for forecasting the load profile on the national grid, then comparing the results and analyzing estimation errors. Auto Regressive Integrated Moving Average Model (ARIMA) is used as a benchmark, to complement the predictions of the ANN method and make it stronger. The research results show that the estimation of the load profile of the National Interconnected System can be automated with acceptable accuracy, thus providing a tool for decision support power procurement officers responsible and meet the minimum requirements necessary to provide reasonable assurance suitable for load dispatch.
Keywords : Load Profile, Artificial Intelligence; Artificial Neural Network; ARIMA; Load Dispatch.