Total Power Quality Index for Electrical Networks Using Neural Networks

The aim of this paper is to present a Total Power Quality Index (TPQI) for estimating Power Quality (PQ) at a specific site of the electric network using Artificial Neural Network (ANN) methodology techniques and a variety of methods from Machine Learning. The main advantage of this methodology is the avoidance of certain arbitrary parameters used in previous research efforts implementing a Fuzzy Index Model. For the purposes of this study and for the training process of the proposed ANN model additional data derived from Power Quality measurements along with an appropriate questionnaire which gathered the opinion of experts was used. The proposed methodology appeared to be flexible, reprogrammable with the capability to assimilate new measurements and estimations in any specific case study. The outcome of the proposed ANN model is compared with the corresponding previously introduced Fuzzy Index model over three different case studies, providing significant results for further improvement. (C) 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).