Intelligent System for Predicting the Consumption of Mobile Electricity consumption
Intelligent System for Predicting the Consumption of Mobile Electricity consumption
- Tez No: 870647
- Danışmanlar: PROF. DR. SEFER KURNAZ
- Tez Türü: Diş Hekimliği Uzmanlık
- Konular: Elektrik ve Elektronik Mühendisliği, Electrical and Electronics Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2022
- Dil: İngilizce
- Üniversite: Altınbaş Üniversitesi
- Enstitü: Lisansüstü Eğitim Enstitüsü
- Ana Bilim Dalı: Elektrik Ana Bilim Dalı
- Bilim Dalı: Elektrik Eğitimi Bilim Dalı
- Sayfa Sayısı: Belirtilmemiş.
Özet
The electricity is an important source of energy in human society. And the forecasting of electricity consumption has become one of the significant essential domains of investigation in the modern ages. Actual energy requirement forecasting is critical in power system services and planning. In this thesis, we will describe how to build business model to choose a perfect Method to predict electrical energy demand in the Future depending on its behavior and Other Parameter like temperature and developing enhanced Clustering methods for smart meters depend on behavior not only pattern with name Sequence clustering. In this thesis, we build clustering for smart meter depend on its daily behavior not only some characteristics like min, max and avg. Eventually, first two years' data is used for training the models and to obtain prediction data for third year. Results demonstrate that hybrid techniques perform better than all other three techniques accuracy 99% whereas SVM, naïve Bayes and decision tree have accuracy of 93%, 93% and 98%, respectively.
Özet (Çeviri)
The electricity is an important source of energy in human society. And the forecasting of electricity consumption has become one of the significant essential domains of investigation in the modern ages. Actual energy requirement forecasting is critical in power system services and planning. In this thesis, we will describe how to build business model to choose a perfect Method to predict electrical energy demand in the Future depending on its behavior and Other Parameter like temperature and developing enhanced Clustering methods for smart meters depend on behavior not only pattern with name Sequence clustering. In this thesis, we build clustering for smart meter depend on its daily behavior not only some characteristics like min, max and avg. Eventually, first two years' data is used for training the models and to obtain prediction data for third year. Results demonstrate that hybrid techniques perform better than all other three techniques accuracy 99% whereas SVM, naïve Bayes and decision tree have accuracy of 93%, 93% and 98%, respectively.
Benzer Tezler
- Konteyner liman operasyonlarının makine öğrenmesi yöntemleri ile analizi
Analysis of container port operations using machine learning methods
ÜSTÜN ATAK
Doktora
Türkçe
2022
İstanbul Teknik ÜniversitesiDeniz Ulaştırma İşletme Mühendisliği Ana Bilim Dalı
Yükleniyor...
- Energy efficient base station deployment in ris-assisted next-generation hetnets
Rıs-destekli yeni nesil heterojen ağlarda enerji verimli baz istasyonu yerleşimi
MİNE ARDANUÇ
- Veri madenciliği ile geliştirilen bir akıllı buzdolabı ve market sepet analizi sistemi
A smart refrigerator developed by data mining and market basket analysis system
ENGİN OĞUZAY
- Spor organizasyonlarında dijital teknolojilerle entegre sıfır atık uygulamaları ve yenilikçi çözümler
Innovative solutions and zero waste applications integrated with digital technologies in sports organizations
ZEYNEP BOYAMA
Yüksek Lisans
Türkçe
2026
İstanbul Üniversitesi-CerrahpaşaSpor Yönetim Bilimleri Ana Bilim Dalı
Yükleniyor...
- Yüksek binalarda asansörlerin tasarımı ve değerlendirilmesi için bir uzman sistem
An Expert system for the design and evaluation of the elevators in high buildings
NURAY ÇANKAYA
