Ubicación óptima de transformadores de distribución, considerando restricciones físicas existentes mediante técnicas de conglomerados combinadas con la teoría de grafos

Abstract

This work focuses on the location and sizing optimization of distribution transformers as an alternative to what is currently happening in local distribution companies, where transformers are located with simplified procedures that omit relevant criteria, such as losses and total cost minimization of the implemented systems.

A methodology which minimizes the system costs, taking into account physical restrictions and road topology, was developed in order to determine the number, capacity and location of transformers in a distribution network. To do this, three stages were developed. First, a clustering algorithm that considers electrical demand momentums, instead of only considering distances like certain traditional models, was developed, which allowed finding the location of the transformers. Then, the transformers’ capacities were determined according to the optimal users’ clustering subject to a constraint of maximum distance, calculated between each transformer and its furthest user, which guaranteed that the voltage drop met the established limits. Finally, the optimal number of transformers was determined by minimizing the total system annualized cost.