Análisis técnico-económico para la implementación óptima de sistemas de medición inteligente en el área de concesión de la CENTROSUR empleando un algoritmo de inteligencia artificial

Abstract

This paper analyzes the implementation of the smart metering system in the concession area of the Empresa Eléctrica Regional Centro Sur CA, using a clustering algorithm to select where the clients that meet the appropriate technical and economic conditions are located for the installation of smart metering equipment. The places where smart metering systems will be installed, must be selected in compliance with parameters related to energy consumption, non-technical losses, invoiced values for energy consumption, customer stratification and debt values for energy consumption. For this, an unsupervised learning method will be used through the k-means artificial intelligence algorithm, based on the data from the Centrosur customer registry, classifying customers who meet the indicated characteristics, without considering any type of criteria based in the intuition or in the experience of the operators of the Centrosur system. The implementation of this type of technology, in addition to solving problems by reading meters, cutting and reconnecting the electric power service remotely, control of losses and energy balances, will allow having a real-time data record of energy consumption of the client, transformers and medium voltage branches. Additionally, it will be possible to know the current state of the parameters of the distribution system, and this information can be used in the future to adequately manage the planning of the repowering or expansion of the electrical system, and the possible connection of distributed generation systems