Research Repository

An IoT-based Prediction Technique for Efficient Energy Consumption in Buildings

Goudarzi, Shidrokh and Anisi, Mohammad Hossein and Soleymani, Seyed Ahmad and Ayob, Masri and Zeadally, Sherali (2021) 'An IoT-based Prediction Technique for Efficient Energy Consumption in Buildings.' IEEE Transactions on Green Communications and Networking. ISSN 2473-2400

[img]
Preview
Text
TGCN.pdf - Accepted Version

Download (1MB) | Preview

Abstract

Today, there is a crucial need for precise monitoring and prediction of energy consumption at the building level using the latest technologies including Internet of Things (IoT) and data analytics to determine and enhance energy usage. Data-driven models could be used for energy consumption prediction. However, due to high non-linearity between the inputs and outputs of energy consumption prediction models, these models need improvement in terms of accuracy and robustness. Therefore, this work aims to predict energy usage for the optimum outline of building-extensive energy distribution strategies based on a lightweight IoT monitoring framework. To calculate accurate energy consumption, an enhanced hybrid model was developed based on Auto-Regressive Integrated Moving Average (ARIMA) and Imperialist Competitive Algorithm (ICA). The parameters of the ARIMA model were optimized by adapting the ICA technique that improved fitting accuracy while preventing over-fitting on the acquired data. Then, Exponentially Weighted Moving Average (EWMA) was applied to monitor the predicted values. The proposed AIK-EWMA hybrid model was assessed based on the actual power consumption data and validated using mathematical tests. As compared to previous works, the findings revealed that the hybrid model could accurately predict power consumption for green building automation applications.

Item Type: Article
Uncontrolled Keywords: Artificial intelligence, auto-regressive integrated moving average, imperialist competitive algorithm, building energy consumption, prediction
Divisions: Faculty of Science and Health > Computer Science and Electronic Engineering, School of
Depositing User: Elements
Date Deposited: 28 Jun 2021 14:19
Last Modified: 28 Jun 2021 15:15
URI: http://repository.essex.ac.uk/id/eprint/30663

Actions (login required)

View Item View Item