Deep Learning-Based Energy Consumption Prediction in IOT Networks
International Journal of Drug Delivery Technology · 29.06.2026 · 16 (59s)
DOI: 10.25258/ijddt.16.59s.95 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 29.06.2026. Verfasst von Vivek Kumar Tamta und 2 weiteren Autor:innen.
Abstract (Originalsprache)
In the era of mass proliferation of the Internet of Things (IoT), the smart home has evolved to rely heavily on connected sensing, communication and automated control of diverse devices. However, the proliferation of smart appliances, sensors, and intelligent control systems leads to significant increases in the energy use of households, which presents significant challenges for efficient energy use, demand-side optimization and sustainable operation of smart grids. Precise forecasting of energy usage is crucial for the energy scheduling, load balancing, anomaly detection, and energy resource management in smart home IoT systems. The existing forecasting techniques such as statistical and conventional machine learning models struggle to capture the complex energy usage patterns of dynamic IoT environments, especially when the data is nonlinear and temporally dependent with high dimensions. To overcome these drawbacks, this study suggests an Optimized Attention-Based Hybrid Convolutional Neural Network – Bidirectional Long Short-Term Memory (CNN-BiLSTM) network for energy consumption forecasting in smart home IoT networks. The proposed framework combines the feature extraction ability of the Convolutional Neural Networks (CNNs) with the bidirectional temporal learning capability of the Bidirectional Long Short-Term Memory (BiLSTM) networks to effectively model complex energy usage behaviors. In addition, there is an attention mechanism that dynamically assigns importance weights to influential temporal features, thus enhancing the precision of prediction and the representation of the temporal features. A thorough data preprocessing methodology is integrated to tackle data sparsity, normalization, temporal feature engineering, and sequence creation via sliding windows, ensuring robust learning and the model's effectiveness. Multiple energy forecasting scenarios are employed to assess the proposed model, based on the publicly-accessible smart-home energy consumption dataset. Very popular performance measures are used in the experimental analysis: Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Coefficient of Determination (R²). The comparative experiments are conducted with other methods such as Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) and CNNLSTM architectures. The results show the proposed Attention-based CNN-BiLSTM model outperforms the existing state-of-the-art approaches in terms of prediction accuracy, temporal learning capability, and generalization capability.
Zitieren
APA
Tamta, V. K., Kumar, J., & Pal, J. (2026). Deep Learning-Based Energy Consumption Prediction in IOT Networks. International Journal of Drug Delivery Technology, 16(59s). https://doi.org/10.25258/ijddt.16.59s.95
Vancouver
Tamta VK, Kumar J, Pal J. Deep Learning-Based Energy Consumption Prediction in IOT Networks. International Journal of Drug Delivery Technology. 2026;16(59s). doi:10.25258/ijddt.16.59s.95
BibTeX
@article{Tamta2026,
author = {Tamta, Vivek Kumar and Kumar, Jogendra and Pal, Jayant},
title = {Deep Learning-Based Energy Consumption Prediction in IOT Networks},
journal = {International Journal of Drug Delivery Technology},
year = {2026},
volume = {16},
number = {59s},
doi = {10.25258/ijddt.16.59s.95}
}
EUVIMED