Quatérnios: estrutura algébrica e aplicação em análise de sinais
Data
Autor(es)
Título da Revista
ISSN da Revista
Título de Volume
Editor
Abstract
This work presents a study on quaternion algebra and its application to the analysis of electrical signals, proposing a computational implementation of the Quaternion Fourier Transform (QFT) as an alternative to the classical Fast Fourier Transform (FFT). Quaternions were introduced by William Rowan Hamilton in 1843 as a three-dimensional extension of complex numbers, overcoming the limitations of “triplets” and establishing the foundation for operations in four dimensions. The theoretical framework covers the algebraic properties of quaternions, including addition, multiplication, modulus, norm, conjugation, and polar representation, as well as the definitions of unit and pure quaternions and the quaternionic exponential and logarithmic functions. Based on these formulations, the mathematical structure required to understand the QFT is developed, highlighting the structural differences in relation to the complex transform, especially regarding multiplicative non-commutativity and the existence of left- and right-sided versions. In the practical application, the study examines harmonic distortions in electric power systems caused by nonlinear loads such as inverters, rectifiers, electric arc furnaces, and others. These distortions are traditionally evaluated using the FFT, but this work aims to employ the QFT for processing three-phase signals with respect to integer harmonics, interharmonics, and subharmonics. The methodology consisted of implementing the QFT computationally in Python, based on the mathematical principles established throughout the work, and comparing its results with those of the traditional FFT. This implementation sought to assess the numerical accuracy, efficiency, and interpretative potential of the QFT in the context of electrical signal analysis. The results indicate that the QFT provides a promising alternative to the classical transform, enabling more comprehensive analyses and more detailed representations of three-dimensional signals, although at the cost of increased mathematical and computational complexity.
