Evaluation of Harmonic Contribution Impacts in the Electric Grid Through Linear Regression, Artificial Neural Networks and Regression tree

This work shows the evaluation of the harmonic contribution at the common coupling point (CCP) of the electric network of the Federal University of Pará (UFPA), which connects four main feeders that have linear and nonlinear loads connected along them. In this article, emphasis is placed on the CCP with the local electric utility and the four electric power feeders of the campus, in order to evaluate the harmonic contribution of each feeder in the CCP of the university, using linear regression techniques and computational intelligences such as artificial neural networks and regression trees. The results of the three analyzes are compared to each other, in order to classify the feeders in relation to their respective impact on the campus electrical grid. The analysis results show that one of the feeders has a more significant impact on the voltage distortion at the CCP of the university, giving subsidies for a more efficient mitigating action.