Algoritmo de Aprendizaje-Q aplicado al control de la calidad de potencia en redes de media tensión con generación distribuida

Abstract

In this work the implementation of dynamic programming is studied as control logic of the power quality in medium voltage electric distribution systems with distributed generation. This was done with the aim of optimizing the power quality indicators and the operation costs; and also to improve the energy efficiency, as well as to take advantage of the transport capacity and of the distribution of the network. With this purpose the work was carried out on a real distribution system belonging to the EPEC company of the Córdoba province in Argentina. The real system behavior was modeled by means of the software ATP/EMTP. Afterwards, a platform was built for the control using MATLAB and an algorithm was developed to process the information and to adjust the system under the new operation conditions. The results indicate that it is possible to control, to regulate and to improve quality indicators, optimizing the distributed generation entrance.