Enhancing e-governance in the ministry of electricity in Iraq using artificial intelligence
Irak elektrik bakanlığında e-yönetişimin yapay zeka kullanılarak iyileştirilmesi
- Tez No: 788139
- Danışmanlar: PROF. DR. NİZAMETTİN AYDIN
- Tez Türü: Yüksek Lisans
- Konular: Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol, Computer Engineering and Computer Science and Control
- Anahtar Kelimeler: ARIMA, e-governance, energy consumption, long-term electricity demand forecasting, LSTM, ARIMA, e-governance, energy consumption, long-term electricity demand forecasting, LSTM
- Yıl: 2023
- Dil: İngilizce
- Üniversite: Yıldız Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Bağımlılık Ana Bilim Dalı
- Bilim Dalı: Bilgisayar Bilimi ve Mühendisliği Bilim Dalı
- Sayfa Sayısı: 98
Özet
Accurately estimating energy consumption over long periods is critical for companies that distribute and supply electricity, whether they work for the government or the private sector. It is essential to increase the quality of future energy production, especially in countries like Iraq that have been experiencing energy crises for a long time. It is possible to both improve gaps in energy resource needs and accurately estimate energy consumption by implementing artificial intelligence within an electronic governance system. Such a system can also help ensure that specific areas receive the appropriate energy supply by monitoring existing stations and identifying which stations need to be expanded. This determination is possible with accurate reports sent to the relevant authorities at regular intervals. In this study, the two-year electricity consumption data of the Iraqi Ministry of Electricity for the city of Mosul between the years 2020-2021 was used, taking into account the effect of external factors, such as the effect of weather conditions. Various models have been proposed, such as the Exogenous (SARIMax) model, which includes the use of neural networks, and statistical models, such as Long-Short-Term Memory (LSTM), Stacked LSTM, Bidirectional LSTM, Auto-Regression Integrated Moving Average (ARIMA) and Seasonal Auto, and Regressive Integrated Moving Average. To predict future electricity consumption and improve the quality of power generation, the models were trained with two categories of data: with or without weather data. The models were trained over a series of epochs (10, 20, and 30) with three different optimizers for each stack of these epochs. While the Bi-LSTM model had the best results in the experiments, with results of 9.207 RMSE, the statistical models ARIMA and SARIMAx reached RMSE values of 14.610 and 15.743, respectively.
Özet (Çeviri)
Accurately estimating energy consumption over long periods is critical for companies that distribute and supply electricity, whether they work for the government or the private sector. It is essential to increase the quality of future energy production, especially in countries like Iraq that have been experiencing energy crises for a long time. It is possible to both improve gaps in energy resource needs and accurately estimate energy consumption by implementing artificial intelligence within an electronic governance system. Such a system can also help ensure that specific areas receive the appropriate energy supply by monitoring existing stations and identifying which stations need to be expanded. This determination is possible with accurate reports sent to the relevant authorities at regular intervals. In this study, the two-year electricity consumption data of the Iraqi Ministry of Electricity for the city of Mosul between the years 2020-2021 was used, taking into account the effect of external factors, such as the effect of weather conditions. Various models have been proposed, such as the Exogenous (SARIMax) model, which includes the use of neural networks, and statistical models, such as Long-Short-Term Memory (LSTM), Stacked LSTM, Bidirectional LSTM, Auto-Regression Integrated Moving Average (ARIMA) and Seasonal Auto, and Regressive Integrated Moving Average. To predict future electricity consumption and improve the quality of power generation, the models were trained with two categories of data: with or without weather data. The models were trained over a series of epochs (10, 20, and 30) with three different optimizers for each stack of these epochs. While the Bi-LSTM model had the best results in the experiments, with results of 9.207 RMSE, the statistical models ARIMA and SARIMAx reached RMSE values of 14.610 and 15.743, respectively.
Benzer Tezler
- The role and impact of e-governance in enhancing human resources management: Study Case of Babylon University
İnsan kaynakları yönetiminin geliştirilmesinde e-yönetişimin rolü ve etkisi: Babylon Üniversitesi Çalışma Örneği
MOHAMMED FARHAN KZAR
Yüksek Lisans
İngilizce
2022
İşletmeİstanbul Gelişim Üniversitesiİşletme Ana Bilim Dalı
PROF. DR. ANTON ABDULBASAH KAMIL
- Siyasi otoritenin olmadığı bölgelerde mülkiyet
Property in areas without political authority
CEREN CİHAN SOYSAL
Doktora
Türkçe
2025
Jeodezi ve Fotogrametriİstanbul Teknik ÜniversitesiGeomatik Mühendisliği Ana Bilim Dalı
PROF. DR. HASAN HAKAN YAVAŞOĞLU
- Ar-Ge faaliyetlerinin firmaların finansal performansına olan etkileri üzerine bir araştırma
A research on the effects of R&D activities on the financial performance of firms
PELİN LATİFOĞLU ÇİFTÇİ
- Water resource management in libya from the farmers perspective
Çiftçi perspektifinden libya'da su kaynakları yönetimi
ALI KHALEEFA ALI YOUNES
Yüksek Lisans
İngilizce
2021
İnşaat MühendisliğiKarabük Üniversitesiİnşaat Mühendisliği Ana Bilim Dalı
DR. ÖĞR. ÜYESİ FATİH SAKA
- Türkiye'de akıllı kent uygulamalarının afet yönetimi boyutuyla analizi
Analysis of smart city applications in Türkiye with the dimension of disaster management
MEHTAP KILIÇ
Doktora
Türkçe
2025
Kamu YönetimiDokuz Eylül ÜniversitesiAfet Yönetimi Ana Bilim Dalı
PROF. DR. YUNUS EMRE ÖZER