International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 11 | Nov 2024
p-ISSN: 2395-0072
www.irjet.net
Application of Machine and Deep Learning Models in Smart Grid Functionalities: A Survey Abdulaziz Salihu Aliero1, Neha Malhotra2 1,2School of Computer Applications, Lovely Professional University, Punjab, India
---------------------------------------------------------------------***--------------------------------------------------------------------
Abstract - These days, one of the necessities for humankind
energy use and lower demand during peak hours, thus avoiding power overload [2].
is electricity. The idea of smart grids was developed to solve problems and difficulties in the traditional grid’s ability to transmit electricity. The survey looks at relevant literature on machine learning (ML) and deep learning (DL) for smart grid applications, with a focus on works published in the last three years. A variety of databases, including Web of Science, Scopus, IEEE Xplore, Science Direct, and Google Scholar, were used to gather research publications. The research on machine learning (ML) and deep learning (DL) approaches used for load forecasting, grid stability, load optimisation, and anomaly detection in smart grids is systematically reviewed in this survey. Additionally, it offers more research problems for using DL and ML technology to make genuinely intelligent grids a reality. The survey will assist the industry and researcher in their future study and analysis of the latest advancements in smart grid technology.
Machine learning (ML) is the prediction and ongoing learning from available data. ML is composed of various algorithms that evaluate the data and generate conclusions or forecasts about the current situation. Deep Learning is a machine learning subfield that focuses on artificial neural network techniques modeled after the brain’s architecture and operations [3]. The integration of ML and DL has various applications in the context of smart grids. These applications include load forecasting, grid stability, energy optimization analysis, anomaly detection, and ensuring the stability of the Smart Grid [4]. Furthermore, deep learning and machine learning are utilized in energy fore- casting to evaluate vast volumes of data and produce precise forecasts. These methods are capable of handling substantial volumes of data and extracting important traits to increase predicting accuracy [5]. Hossain et al. carried out a thorough investigation of the use of machine learning in smart grids, highlighting its potential in load forecasting, data acquisition, and stability assessment [6]. They emphasized the need for advanced techniques to handle the massive data volume generated by smart grids. Where by Massaoudi et al. reviewed the advancements and prospects of deep learning in smart grids, discussing its applications in energy forecasting, fault detection, and cybersecurity awareness [7]. Deep learning, in particular, involves stacking and connecting different learning layers that accurately map the link between input data and output [8].
Key Words: Smart Grid, Machine Learning, Deep Learning, Models
1. INTRODUCTION A smart grid is an interconnected power system that uses automation and digital communication technologies to increase the long-term viability, dependability, and effectiveness of the production, dispersion, and use of electricity. It makes it possible for the Electricity Company and customers to communicate in both directions, which permits real-time monitoring and management of electricity consumption. Smart grids collect data on electrical supply and demand using sensors, meters, and other devices. This allows for improved resource management and grid stability through energy optimization, distribution, and reducing transmission defeats. Smart grids can contribute to energy efficiency and savings on costs for consumers and utilities. In addition to facilitating demand- side management and load balancing, the deployment of smart grids allows demand response programs, in which users can modify their electricity consumption according to price indications or system circumstances. Overall, smart grids are essential in modernizing the electrical sector and addressing the challenges of climate change, rising energy prices, and the need for a more sustainable and resilient electricity infrastructure [1]. The smart grid provides various services in real life that enable demand management, allowing clients and proprietors to control
© 2024, IRJET
|
Impact Factor value: 8.315
This paper primarily focuses on understanding the opportunities for using the application of machine and deep learning models in smart grids. This will involve a detailed discussion of the various techniques, algorithms, and models used in machine deep learning for the smart grid. The paper will also enumerate the different data sources that were used in machine and deep learning for understanding how the smart grid works, like communication networks, intelligent meters, detectors, grid management networks, solar and wind turbines, grid storage systems, and data management. In addition, the paper will discuss why smart grid are important, and the architecture of smart grid. Finally, this survey comes to a close with a discussion of where research in this area will go in the future. As part of this, it will be necessary to identify research areas that need additional study, such as the creation of deep learning algorithms that are more accurate
|
ISO 9001:2008 Certified Journal
|
Page 679