|
Evaluating high-consumption and unusual subscribers in the smart gas meter network using machine learning and the Internet of Things in the cloud environment
|
K. Aghaei Badr   , H. Mehrmanesh , A. Fadavi Asghari |
| Department of Industrial Management, CT.C., Islamic Azad University, Tehran, Iran |
|
|
Abstract: (20 Views) |
With the advancement of new technologies, the use of smart gas meters as a tool for managing energy consumption and optimizing energy resources is expanding. These meters can collect and analyze consumption data in real time with the help of Internet of Things (IoT) and machine learning. The aim of this research is to evaluate and identify high-consumption and abnormal subscribers in the smart gas meter network using machine learning algorithms and Internet of Things technology in a cloud environment. This research seeks to provide solutions to improve energy consumption management and reduce costs by identifying abnormal consumption patterns and providing optimization suggestions to subscribers. The importance of energy consumption management and the implementation of related policies have required governments to identify high-consumption subscribers and separate them from low-consumption subscribers. Accordingly, policies are being developed to fine or punish high-consumption subscribers based on their consumption and even reward low-consumption subscribers. This is possible more efficiently using a smart meter network in which data is transferred in real time on the Internet of Things network and stored in a cloud computing environment.
In this research, in line with this policy, an attempt has been made to design a model to identify and control high-consumption and irregular subscribers in the smart gas meter network. This model includes 5 variables: annual consumption, monthly consumption, consumption period, household size, and subscription type, which were implemented using 4 machine learning algorithms: random forest, decision tree, nearest neighbor, and XG boost.
The results show that the random forest algorithm was able to classify and identify high-use subscribers with 92% accuracy, followed by the XG boost algorithm with 91% accuracy, and then the nearest neighbor and decision tree algorithms with 90% accuracy.
The conclusion of this research shows that the use of machine learning algorithms and IoT technology in the smart gas meter network can help to accurately identify high-consumption and abnormal subscribers. This not only leads to energy consumption optimization and cost reduction, but also enables the implementation of effective policies for energy consumption management. |
|
| Keywords: High-Consumption Subscribers, Aberrant Subscribers, Meter Network, Gas, Machine Learning, Internet of Things, Cloud Computing |
|
|
Full-Text [PDF 843 kb]
(5 Downloads)
|
Type of Study: Research |
Subject:
General
Received: 2025/03/26 | Accepted: 2025/06/12 | Published: 2025/07/1
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
