Comparative Analysis of Electricity Consumption Forecast

This study aims to make a comparative analysis of electricity consumption forecast using artificial intelligence (AI) and statistical models. In order to reduce the current deficits of countries, it is of great importance to predict the future electricity consumption amount and plan the power plant...

Full description

Saved in:
Bibliographic Details
Main Authors: Mehet Ali ARSLAN, Tarık TALAN
Format: Article
Language:English
Published: Bursa Technical University 2025-06-01
Series:Journal of Innovative Science and Engineering
Subjects:
Online Access:http://jise.btu.edu.tr/en/download/article-file/4520562
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849245461945253888
author Mehet Ali ARSLAN
Tarık TALAN
author_facet Mehet Ali ARSLAN
Tarık TALAN
author_sort Mehet Ali ARSLAN
collection DOAJ
description This study aims to make a comparative analysis of electricity consumption forecast using artificial intelligence (AI) and statistical models. In order to reduce the current deficits of countries, it is of great importance to predict the future electricity consumption amount and plan the power plant capacities accordingly. Electricity is an energy source that is extremely difficult to store when used in sectors such as industry and housing. Therefore, the electricity produced must be consumed immediately without causing energy losses and waste. In this context, ensuring the balance between electricity production and consumption can correctly contribute to the management of the current deficit by increasing economic efficiency. In the current study, Türkiye's hourly electricity consumption data between 2016 and 2024 were examined. These data were transformed into a 108-month consumption data set. Seven different models, namely Auto-ARIMA, Holt-Winters, Theta, ETS, TBATS, NNETAR and MLP, were used in the analyses. Among the models, NNETAR and MLP are AI based, and the others are statistical-based models. In this way, the effectiveness of different model types in electricity consumption estimations was compared. In this study, the Auto-ARIMA model stood out with a 3.77% MAPE error rate. When such studies are considered within the framework of countries' energy policies, they can make a significant contribution to reducing the current deficit of the country's economy. As a result of the study, it was concluded that the Auto-ARIMA model should be taken into consideration when making estimates on how many Megawatt power plants should be built in order to meet future energy needs in shaping energy policies in Türkiye.
format Article
id doaj-art-7192ad59dfe74f04a58de9027ba325fb
institution Kabale University
issn 2602-4217
language English
publishDate 2025-06-01
publisher Bursa Technical University
record_format Article
series Journal of Innovative Science and Engineering
spelling doaj-art-7192ad59dfe74f04a58de9027ba325fb2025-08-20T03:58:49ZengBursa Technical UniversityJournal of Innovative Science and Engineering2602-42172025-06-01918910210.38088/jise.1619782Comparative Analysis of Electricity Consumption ForecastMehet Ali ARSLAN0https://orcid.org/0000-0002-0701-059XTarık TALAN1https://orcid.org/0000-0002-5371-4520Gaziantep Islam Science and Technology UniversityGaziantep Islam Science and Technology UniversityThis study aims to make a comparative analysis of electricity consumption forecast using artificial intelligence (AI) and statistical models. In order to reduce the current deficits of countries, it is of great importance to predict the future electricity consumption amount and plan the power plant capacities accordingly. Electricity is an energy source that is extremely difficult to store when used in sectors such as industry and housing. Therefore, the electricity produced must be consumed immediately without causing energy losses and waste. In this context, ensuring the balance between electricity production and consumption can correctly contribute to the management of the current deficit by increasing economic efficiency. In the current study, Türkiye's hourly electricity consumption data between 2016 and 2024 were examined. These data were transformed into a 108-month consumption data set. Seven different models, namely Auto-ARIMA, Holt-Winters, Theta, ETS, TBATS, NNETAR and MLP, were used in the analyses. Among the models, NNETAR and MLP are AI based, and the others are statistical-based models. In this way, the effectiveness of different model types in electricity consumption estimations was compared. In this study, the Auto-ARIMA model stood out with a 3.77% MAPE error rate. When such studies are considered within the framework of countries' energy policies, they can make a significant contribution to reducing the current deficit of the country's economy. As a result of the study, it was concluded that the Auto-ARIMA model should be taken into consideration when making estimates on how many Megawatt power plants should be built in order to meet future energy needs in shaping energy policies in Türkiye.http://jise.btu.edu.tr/en/download/article-file/4520562electrical energyelectricity production forecastartificial intelligencedeep learning
spellingShingle Mehet Ali ARSLAN
Tarık TALAN
Comparative Analysis of Electricity Consumption Forecast
Journal of Innovative Science and Engineering
electrical energy
electricity production forecast
artificial intelligence
deep learning
title Comparative Analysis of Electricity Consumption Forecast
title_full Comparative Analysis of Electricity Consumption Forecast
title_fullStr Comparative Analysis of Electricity Consumption Forecast
title_full_unstemmed Comparative Analysis of Electricity Consumption Forecast
title_short Comparative Analysis of Electricity Consumption Forecast
title_sort comparative analysis of electricity consumption forecast
topic electrical energy
electricity production forecast
artificial intelligence
deep learning
url http://jise.btu.edu.tr/en/download/article-file/4520562
work_keys_str_mv AT mehetaliarslan comparativeanalysisofelectricityconsumptionforecast
AT tarıktalan comparativeanalysisofelectricityconsumptionforecast