Evaluation and Analysis of Least Square and Exponential Regression Techniques for Consumer Energy Consumption Requirement

Abstract

The aim of this paper is to conduct the analysis of the load forecast and energy demand. The least-square regression and exponential regression models were used in this study to anticipate long-term power demand for a twenty-year (20 years) projection in the Nigerian power system. Residential load demand, commercial load demand, and industrial load demand are plotted in the model's Matlab platform implementation. The findings indicate that the energy produced by each generating station, including Sapele Thermal Power Station, Egbin Thermal Power Station in Lagos, etc., is woefully insufficient. Additionally, the outcome reveals that there is a discrepancy between the expected energy demand (MW) and the available power (or capacity allocated). Evidently, the comparison plot for the linear and exponential models, which exhibit a similar pattern of prediction (the least-square exhibits linear behaviour, while the exponential displays non-linear behaviour), shows that the linear model provides results that are more accurate than the exponential.