SUMMARYThis work presents an alternative approach using the differential logic associated to artificial neural networks (ANNs) in order to distinguish between inrush currents and internal faults for the protection of power transformers. The radius basis function (RBF) neural network is proposed as an alternative approach in order to distinguish the situations described, using a smaller amount of data for training purposes, in some cases, if compared with networks such as the multi-layer perceptron (MLP). The ANN results are then compared to those obtained by the traditional differential protection algorithm. An ANN approach for correction of saturated current signals is also presented.
Current transformers (CTs) are present in ElectricPower Systems for protection and measurement purposes and they are susceptible to the saturation phenomenon. This paper presents an alternative approach to the correction of distorted waveforms caused by CT saturation. The method uses Recurrent Artificial Neural Networks (ANN) algorithms. As an example of an application, a complete protection system for a power transformer based on the deferential logic has been utilized. The EMTP-ATP software has been chosen as the computational tool to simulate the electrical system in order to generate data to train and test the ANNs. Many ANN architectures were trained and tested. Encouraging results related to the application of the new method are presented.
Este trabalho apresenta um sistema completo de proteção diferencial para transformadores de potência, através da teoria de Redes Neurais Artificiais (RNAs). O método proposto trata a classificação do sistema de proteção como um problema de reconhecimento de padrões e constitui um método alternativo aos algoritmos convencionais. Muitos fatores, tais como a energização do transformador e a saturação dos TCs, podem causar uma operação inadequada do relé de proteção. Um sistema de proteção completo foi desenvolvido, incluindo um módulo baseado em RNA em substituição aos filtros harmônicos, usados no algoritmo convencional. Este módulo se constituiu de uma RNA tipo MLP Backpropagation para a classificação de sinais. Abordagens baseadas na reconstrução dos sinais distorcidos causados pela saturação dos TCs são também propostas. Essa análise foi realizada através do emprego de RNAs Recorrentes de Elman, utilizadas para reconstruir os sinais distorcidos pela saturação dos TCs. Essas rotinas foram adicionadas ao algoritmo final de proteção. O desempenho dos algoritmos propostos foi comparado ao do algoritmo convencional de proteção de transformadores, em termos de velocidade e precisão de resposta. Com a utilização de uma ferramenta de inteligência artificial em um algoritmo completo de proteção de transformadores, uma solução precisa, rápida e eficiente foi obtida, se comparada aos métodos convencionais. This paper presents a complete differential protection system for power transformers, applying the Artificial Neural Network (ANN) theory. The proposed approach treat the classification of the protection system as a problem of pattern recognition and as an alternative method to the conventional algorithms. Several factors such as, for example, transformer energization and CT saturation can cause an inadequate operation of the protection relay. A complete protection system was developed, including an ANN-based device in substitution to harmonic filters in use in the conventional algorithm. This stage was carried out by a MLP Backpropagtion ANN to the signals classification. Some approaches concerning the reconstruction of the distorted signals caused by the CTs saturation are also proposed. This analysis was made by Elman recurrent ANNs used to reconstruct the distorted signals caused by CT saturation. These routines are added to the final protection algorithm. With the use of artificial intelligence tools in a complete power transformer protection algorithm, a very precise, fast and efficient solution was obtained, if compared to the conventional methods
RESUMOEste trabalho apresenta uma técnica alternativa para a correção de ondas distorcidas provenientes da saturação dos transformadores de corrente (TCs) através de ferramentas inteligentes baseadas em Redes Neurais Artificiais (RNAs) recorrentes. Os TCs estão presentes em Sistemas Elétricos de Potência com a finalidade de proteção e medição, sendo altamente susceptíveis à saturação. O programa EMTP-ATP (Electromagnetic Transients Program) foi escolhido como ferramenta computacional para a simulação de um sistema elétrico utilizado na geração de dados de treinamento e testes para as RNAs. Muitas arquiteturas de redes neurais artificiais foram treinadas e testadas. Resultados promissores relativos ao novo método são apresentados.
PALAVRAS-CHAVE
ABSTRACTThis paper presents an alternative approach to the correction of distorted waveforms caused by the current transformer (CT) saturation. This method uses the Artificial Neural Networks (ANNs) recurrent algorithms. Current transformers are present in the electric power systems for protection and measurements and they are highly susceptible to the saturation phenomenon. The EMTP-ATP software has been chosen as the computational tool to simulate the electrical system in order to generate data to train and test the ANNs. Many ANN architectures were trained and tested. Encouraging results related to the application of the new method are presented.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.