Typical curve of demand of electric power of the residential, commercial and industrial sector of the city of medellin, using artificial neural networks and algorhythms of interpolation

Abstract

One of the main problems that arise to model the electric power consumption ina certain place, consists on the extraction of the knowledge when this is stored in big volumes of information like, for example, historical registrations. In accordance with this representation, each fact and registration is made up of two components (t, P). t represents the time and P represent the electric power consumed at that instant. The N cases are registered daily; these represent t, P couples.The objective of this study is to map out the variable vector of t and Pperformance been F one of the functions. The model was made with Artificial Neural Network (ANN) which is a Multi Perceptron Layer (MPL). Another form of modeling is using Interpolation Algorithms(AI).