Development of a comparative study between linear regression and decision tree techniques for 3rd harmonic

A comparative study was carried out between the Simple Linear Regression and Decision Tree techniques in the evaluation of the impacts caused by the generation of harmonics, in the distribution network, in a medium voltage substation. Data collection was carried out through a measurement campaign in a 13.8 kV substation that feeds a portion of the companies in the industrial district of the city of Manaus. The objective of this dissertation is to carry out a comparative study of the results of the application of Decision Tree and Simple Linear Regression techniques in the analysis of harmonic impacts in an electrical system. The measurement campaign was carried out for a minimum period of 7 calendar days according to PRODIST. The impact of these harmonics on the electrical system is quite harmful, both for consumers and for utilities. The techniques used in the methodology of this dissertation are the Decision Tree technique, which features the construction of non-parametric models and the Simple Linear Regression technique, which features the construction of parametric models and simple mathematical calculation, easy to interpret and analyze. of data. In this way, this dissertation presented and applied in practice with case studies actions for the analysis of harmonic impacts in electrical energy distribution systems through the construction of mathematical models using Simple Linear Regression analysis and Regression Tree analysis, obtaining important and enlightening results. in the studies carried out, thus validating the techniques applied in the analysis of harmonic impacts in electrical energy distribution systems.