Extração de harmônicos, inter e sub-harmônicos utilizando análise de componentes independentes

• Autor(a) principal: Rufino Júnior, Carlos Antônio
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/35022
• topic: Extração de harmônicos; Extração de inter-harmônicos; Extração de sub-harmônicos; Análise de componentes independentes; Qualidade de energia elétrica; Distúrbios de energia; Harmonics; Extration; Inter-harmonics; Sub-harmonics; Independent components analysis; Electric power quality; Power disturbances; Eletrônica Industrial, Sistemas e Controles Eletrônicos

The increase in demand for electricity, mainly due to the appearance of new electronic equipment and the increase in the number of consumers, has motivated the search for clean and renewable energy sources and the emergence of the context of distributed generation. Distributed generation is characterized by small generation plants connected to the main electrical system. This new context has contributed to the increase of electric power quality problems, with emphasis on the increase of sources of harmonic, inter-harmonic and sub-harmonic components in the Electric Power System. Such components generate a number of problems, such as improper protection operation, measurement errors, line overheating and losses in transformers. In this way, the extraction of harmonic, inter-harmonic and sub-harmonic components of the electric system is of paramount importance. The present work proposes the implementation of the Filters-Analysis of Independent Components (FICA) method adapted to the problem of extraction of harmonics, sub-harmonics and inter-harmonics of the voltage signal and/or current of the electric power system. We evaluated 10 synthetic signals known in the current literature containing harmonics, inter-harmonics and sub-harmonics. In these signals were introduced disturbances such as sagging, oscillatory transients and noise. The results showed that the FICA - adapted method is able to extract the fundamental component and the harmonic, inter-harmonic and sub-harmonic components present in the voltage and/or processed current signal. As metrics, MSE, RRMSE, SD and RSD were used. The results show that the proposed method presents competitive results in view of several advantages over other methods in the literature, since it does not require the clustering step as in methods that use Single Channel Independent Component Analysis (SCICA), as well as method is able to extract components with low energy and components with low and high frequencies. In addition, the proposed method does not require parameters of the monitored electric signal, acts as an adaptive filter, does not suffer from fundamental frequency variations, and has the ability to extract interharmonics and subharmonics adjacent to the harmonic components. For future work it is expected to implement the proposed method online.