Modelo de predicción de demanda del sector residencial de la ciudad de Loja.

This report presents the results of designing a model predicting electricity demand in the residential sector of the Loja city, was taken as a case study the feeder "Calvario", belonging to the substation "San Cayetano", the nature of where users demand feeder was considered as residential. To create the prediction model were considered strictly nonlinear methods based on Artificial Neural Networks (ANN), selecting ANN architecture Nonlinear Auto-regressive, this architecture enables the prediction of time series, relying solely with the number of historical data, which was provided by the Empresa Eléctica Regional de Sur (EERSSA). To get to determine the database that feed entries RNA processing was performed comprehensive database provided, reaching delete data and in some cases made the curve smoothing the data obtained. In the process of testing and validation of the prediction model of electrical demand, it was found that the results obtained by the selected ANN for each prediction fits planted metrics prior to the design process. For the development of the research was done using MatLab software engineering (Matrix Laboratory) used to develop the process of creating predictive model through ANN.

• 1. Verfasser: Cabrera Samaniego, Juan Pablo (author)
• Format: masterThesis
• Sprache: spa
• Veröffentlicht: 2013
• Schlagworte: <DEMANDA DE POTENCIA><REDES NEURONALES ARTIFICIALES>
• Online Zugang: http://dspace.unl.edu.ec/jspui/handle/123456789/13607