Machine learning based smart grid prediction models for enhancing environmental sustainability
Journal of Applied and Natural Science · 21.06.2026 · 18 (2) · 626-640
DOI: 10.31018/jans.v18i2.7443 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied and Natural Science, online erschienen am 21.06.2026. Verfasst von Utpal Barman und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
The growing demand for electricity, driven by population growth and economic expansion, necessitates an efficient power distribution system and environmental sustainability. Smart Grids have emerged as a potential tool for enhancing the stability, reliability and sustainability of the grids. Machine learning models implemented in smart grids can predict consumer demand and improve smart grid performance. In this study, various machine learning models, including Logistic Regression, Support Vector Machines, Decision Trees, Random Forests, Gaussian Naïve Bayes, k-Nearest Neighbours, and Artificial Neural Networks, are used. The results of models with and without Principal Component Analysis integration are highlighted to predict the stability of smart grids. The metrics used in this study to evaluate model performance are accuracy, sensitivity, specificity, F1 score, and AUC-ROC. The results showed that machine learning models without principal component analysis performed better than those with principal component analysis on this dataset. The SVM-RBF kernel model with PCA achieved the highest accuracy of 80.39%, whereas the model without PCA achieved 98.15%. The results show that integrating PCA into ML models does not improve model performance; rather, it decreases accuracy, especially given the dataset’s size and the lower-dimensional feature space of the Smart Grid UCI dataset.
Zitieren
APA
Barman, U., Lahon, P., Kandali, A. B., & Das, D. (2026). Machine learning based smart grid prediction models for enhancing environmental sustainability. Journal of Applied and Natural Science, 18(2), 626-640. https://doi.org/10.31018/jans.v18i2.7443
Vancouver
Barman U, Lahon P, Kandali AB, Das D. Machine learning based smart grid prediction models for enhancing environmental sustainability. Journal of Applied and Natural Science. 2026;18(2):626-640. doi:10.31018/jans.v18i2.7443
BibTeX
@article{Barman2026,
author = {Barman, Utpal and Lahon, Pranobjyoti and Kandali, Aditya Bihar and Das, Dulumani},
title = {Machine learning based smart grid prediction models for enhancing environmental sustainability},
journal = {Journal of Applied and Natural Science},
year = {2026},
volume = {18},
number = {2},
pages = {626--640},
doi = {10.31018/jans.v18i2.7443}
}
EUVIMED