Posicionamento eficiente de chaves em rede de distribuição utilizando algoritmo multiobjetivo com validação de solução por meio de verificação de restrições

• Main Author: Leite, Fabricio da Rocha
• Publication Date: 2019
• Format: Master thesis
• Language: por
• Source: Repositório Institucional da Universidade Federal do Ceará (UFC)
• Download full: http://www.repositorio.ufc.br/handle/riufc/40316
• topic: Engenharia elétrica; Energia elétrica - Otimização; Energia elétrica - Distribuição - Proteção; Localização de falhas; Optimization in electrical systems; Switch placement; Particle swarm optimization

The demand of the consumers for better levels of quality and continuity of the service of energy supply, besides the requirement of regulatory agencies to attend certain quality indicators of this service, has asked of the concessionaires of energy greater efficiency. One of the strategies to solve this problem is the automation of distribution networks, with the installation of intelligent protection equipment, allowing the rapid restoration for temporary faults, or in the case of permanent faults, isolating the faulty part and restoring the unaffected part. This work aims to implement an algorithm that seeks the efficient location of keys in a distribution network, considering criteria of quantity of clients deenergized and number of keys with the capacity constraints of transformers supply, ampacity of the feeders and adequacy level of voltage, in order to contribute to the effective improvement of the reliability of distribution networks. The multiobjective optimization (MOSP) technique, based on a particle swarm (PSO), is based on the paradigm of the interaction between the particles of a swarm (or individuals of a population) that seek a global optimum. The constraints by observing the load of the conductors and transformers, as well as the voltage level at the network nodes are verified through the load flow study by the inverse direct sweep method for all the system's contingencies. For the implementation of the algorithm we used the Python programming language, with representation of the electrical network provided by a data coding supported in graph theory called Node-Depth Representation. The solution developed was applied in the electrical network of the Pici University Campus of the Federal University of Ceará (UFC), whose problems of general power shutdowns are historical, mainly due to the characteristic of being a radial system with only general protection. The equipment installation topology will provide for the subdivision of the campus distribution network, increasing the reliability and availability of the power supply.