Year: 2025
Type: article
Abstract: Efficient energy management requires accurate forecasting of household electricity consumption. This paper presents PowerPredict, a data-driven framework that predicts household energy consumption using machine learning models and historical data [6]. The model utilizes features such as voltage, current intensity, and sub-metering data to forecast global active power. Unlike conventional research that focuses primarily on algorithmic improvements, this work emphasizes practical usability by integrating the trained model into an interactive Streamlit dashboard. The platform enables users to input parameters, receive instant predictions, simulate scenarios, and estimate costs. This study bridges predictive analytics with actionable decision support, providing a simple, IoT-free solution for smarter household energy planning and cost optimization.
Authors: Anmol Singh, Bhoomi Saxena, V. Sathiyasuntharam, Aastha Bansal
Institution: Sharda University
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4416066237