Localización de Distorsiones Armónicas en Sistemas Eléctricos Usando Sensado Comprimido

This article is a study of the state of the art, where the advantages and disadvantages of the different classical digital signal processing techniques used to identify harmonic frequencies that distort the voltage and current waveforms of an electrical network are analyzed. Taking as a reference the classical techniques that determine the harmonic distortion of an electrical signal, an alternative is proposed for the detection of harmonicfrequencies in an electrical waveform, based on compressed sensing algorithms; this technique extracts an efficient and reduced sample of the signal under study, this sample is transformed to the frequency domain by means of a transformation base which, for the present case study is the discrete cosine transform (DCT). The reduced signal in the frequency domain is subjected to a non-linear optimization process, which classifies the coefficients, extracts the most representative ones and makes the least representative ones zero. It is here where the harmonic frequencies, immersed in the electrical waveform come to light in a clear and precise way, with which the harmonic distortion (THD) is calculated. This article shows an example of the compressed sensing technique for the detection of harmonic frequencies in a theoretical sinusoidal signal which is contaminated by third and fifth order harmonics, the optimized coefficient vector is extracted, and the harmonic frequencies are identified.