Affiliations: School of Computing, College of Science, Engineering and Technology, University of South Africa, Pretoria, South Africa
Correspondence:
[*]
Corresponding author: Isaac Osunmakinde, School of Computing, College of Science, Engineering and Technology, University of South Africa, P.O. Box 392, UNISA 0003 Pretoria, South Africa. Tel.: +27 11 670 9104; E-mail: [email protected].
Note: [1] Both authors contributed equally to this article. They have read and approved the final manuscript.
Abstract: Electricity consumption prediction in smart homes and its effective management are global concerns. One of the most important inventions to assist human living, electricity is used by residential users as well as commercial operations. These users often utilize different electronic devices and sometimes consume fluctuating amounts of electricity, generated from smart-grid infrastructure owned by the government or private investors. However, a repeated imbalance is noticeable between the demand and supply of electricity; these disparities are often brought about by different weather profiles such as temperature, wind speed, dew point, humidity and pressure of the electricity consumption locations. Therefore, effective planning through an intelligent data analysis of the electricity load is needed to enable a sustainable distribution among consumers. Such intelligent analysis and planning are activated by the need to visualize the data and predict future electricity consumption within a short period, considering how weather variables affect predictions. Although a variety of compelling state-of-the-art techniques are used for such predictions, they require data engineering improvement for reducing significant predictive errors in short-term load forecasting (STLF). This research deploys a near-zero cooperative probabilistic scenario analysis and decision tree (PSA-DT) model to address the predictive errors facing state-of-the-art models, and analyses the effect each weather profile has on the cooperative model. The PSA-DT is a machine learning (ML) model based on a probabilistic technique (in view of the uncertain nature of electricity consumption), complemented by a DT to reinforce collaboration between the two techniques. Based on detailed experimental intelligent data analytics (IDA) on residential and commercial data loads, together with multiple weather profiles, the PSA-DT model outperforms state-of-the-art models in terms of accuracy to a near-zero error rate. This implies that its deployment for electricity demand in planning smart homes will be of great benefit to various smart-grid operators and homes.
