Intelligent Hybrid Deep Learning Framework for Accurate Energy Consumption Forecasting in IoT-Enabled Homes

View at publisher Summarise with VIRAT
Authors: Premkumar Duraisamy, Deshvanth S, Sanjay M, Gowtham Kumar P
Cited by: 0
References: 22

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Accurate and real-time energy consumption prediction plays a vital role in developing sustainable and intelligent smart homes. This work presents an enhanced hybrid deep learning framework that integrates Convolutional Neural Networks (CNNs), Long Short- Term Memory (LSTM), and Temporal Convolutional Networks (TCNs) to model spatial and temporal dependencies in IoT-based household energy data. The framework leverages multi-source inputs such as smart meter readings, weather parameters, occupancy patterns, and appliance usage to deliver fine-grained energy forecasts. CNN layers extract spatial correlations among appliances, LSTMs capture long-term temporal trends, and TCNs ensure efficient sequence learning with reduced computational cost. Experimental evaluation on real-time IoT datasets demonstrates that the proposed CNN-LSTM- TCN hybrid achieves a prediction accuracy of 97.8%, significantly outperforming baseline models such as Random Forest and standalone LSTM networks. The system also integrates anomaly detection to identify irregular usage behavior and provides actionable insights through a user-friendly energy dashboard. This work contributes to sustainable smart home management by optimizing power utilization, improving forecasting reliability, and promoting energy efficiency in next-generation IoT environments.

Fields of Study & Subjects

Smart Grid Energy Management Energy Load and Power Forecasting Building Energy and Comfort Optimization

Publication details

Journal / source
International Conference on Communication, Computing and Electronics Systems (ICCCES)
Pages
132–137
Publication date
21 January 2026
Language
EN