Power Quality Classification Scheme For A Grid-Integrated Power Distribution System

The energy trajectory has evolved over the years with the inclusion of distributed generation based on renewable energy (RE) technologies on the power system grid. The inclusion of RE technologies improves the power availability and security. However, there are serious technical concern which include voltage and frequency variation. In this paper, a power quality (PQ) classification scheme for a power distribution grid-integrated system with (RE) technologies is proposed. The proposed scheme uses the wavelet packet transform (WPT) for signal processing and feature extraction, and support vector machine (SVM) is used for PQ events classification. The proposed scheme was implemented using the WEKA machine learning platform. Furthermore, the neural network (NN) and naïve bayes (NB) classification techniques ware also tested to validate the proposed technique.

I. Introduction

The evolution of the traditional power generating system has gained acceleration over the past years with the inclusion of distributed generation (DG) being integrated into the existing power grid. The change in power generation trajectory was ignited by environmental impact that fossil fuels emit during power generation. The integration of DGs into the power grid particularly the photovoltaic (PV) and wind energy (WE) systems have been widely used to improve the system reliability and power security. However, the integration of these DG technologies introduces technical constrains such as voltage and frequency variations, which ultimately impacts on the power quality (PQ) of the system [1], [2]. Also, the dependency of the PV and WE technologies on the weather conditions further impacts the energy forecasting techniques, resulting in improper energy management system.