Weibull Distribution Modelling of Wind Energy System and Voltage Total Harmonic Distortion Prediction of Ado-Ekiti Feeder

Abstract

Power quality problems include voltage swell, voltage sag, spikes, harmonic distortion, and the significant obstacles of integrating wind energy systems into the traditional grid network. However, this injection's main goals are to examine the power quality of contemporary power systems and suggest solutions to inadequate power supplies. This paper estimated scale and shape parameters of the Weibull distribution model were 2.7863 and 3.7731, respectively, after ten years of wind data obtained from the National Aeronautic Space Agency. The application of Spline regression was to disaggregate the power metrics that were collected by the aR-6 power quality analyzer. A neural network and linear modeling predicted the voltage total harmonic distortion. The validation of predicted voltage total harmonic distortion was done using an error measure based on a comparison of observed and predicted voltage total harmonic distortion data, as well as a prediction from the neural network and linear modeling. The neural network and linear model have mean square errors of 0.0263 and 0.0381, respectively. The root mean square error, mean average error, and root square for neural network (NN) model prediction were 10.1624, 0.1277, and 0.3774, respectively. The root mean square error, mean average error, and root square for linear model (LM) prediction were 10.1951, 0.1669, and 0.0998, respectively. The predictions were correct since the NN and LM error metrics were quite near, indicating that the actual and anticipated distribution functions were good models.