Projection and analysis of electricity consumption in the water supply system of Bački Vinogradi using historical data and machine learning models (2022-2024)
Year: 2025
Type: conference-paper
Abstract: This study presents the projection and analysis of electricity consumption in the water supply system of Bački Vinogradi, Subotica (2022-2024). Historical data were used for VTA, NTA, VTR, NTR, monthly electricity billing and pumped water volumes (m³). For prediction, Linear and Polynomial Regression, K-Nearest Neighbors, Decision Tree, Random Forest, and ARIMA time series models were applied. Regression and machine learning models enable forecasting of monthly consumption based on the m³ parameter and previous reactive power values, while the ARIMA model uses time series for dynamic forecasting. The analysis includes a comparison of predicted and actual values, identification of the most accurate models, and visualization of results aimed at optimizing consumption.
DOI: 10.5937/vik25345i
Author: Ognjen Ivić
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4414655507