Citas bibligráficas
Orbezo, H., (2011). Análisis estocástico Arima para el modelamiento y predicción de la demanda eléctrica en el sector residencial de Lima Sur [Tesis, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/579
Orbezo, H., Análisis estocástico Arima para el modelamiento y predicción de la demanda eléctrica en el sector residencial de Lima Sur [Tesis]. : Universidad Nacional de Ingeniería; 2011. http://hdl.handle.net/20.500.14076/579
@mastersthesis{renati/703952,
title = "Análisis estocástico Arima para el modelamiento y predicción de la demanda eléctrica en el sector residencial de Lima Sur",
author = "Orbezo Urquizo, Hernán Antonio",
publisher = "Universidad Nacional de Ingeniería",
year = "2011"
}
In the present rosearen applied time series analysis in a stochastic approach ARIMA {Autoregressive Integrated Moving Average), for making forecasts of demand for residential electricity in Lima Sur (variable with irend. seasonality, cycting and randomness), having shown that irnproves efficiency and predictive goodness to compare and validate the resulte obtained with the deterministic prediction Winter technique. An adequate prediction of the energy demand is critical to investment decisions in electricity networks, because the more accurate the predictions lower the risk of incurring unnecessary investments; in that sense, this research represente the fírst attempt to formulate a technical procedure validated to support investment decisions at the level of electrical distribution. As a result of the analysis determined that the residential energy consumption will grow Lima Sur sustained conservative 3,5% on average for the period 2010-2012. Increase also represents the growth of energy consumption at low voltaje, as this energy range has the same behavior as the residential energy consumption for the same area of analysis. The ARIMA methodology provides better predictive results, both graphically and in margin of error and complexity of Ihe variable under study, compared with other methods such as exponential smoothing compound Winter, proving the null hypothesis formutated for the research. to provide results with better efficiency and predictive goodness. So with the ARIMA method results are obtained with lower mean absolute percentage error (1,59%) and lowest sum of squared errors (1359). Keywords: Forecast, residential electricity demand, ARIMA models, time series, stochastic processes, validation.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons