Short-Term Electricity Consumption Forecast Using Datasets of Various Granularities

Arslan, Yusuf; Dilbaz, Aybike Şimşek; Ertekin, Şeyda; Karagöz, Pınar; Birtürk, Ayşenur; Eren, Sinan; Küçük, Dilek

doi:10.1007/978-3-030-04303-2_9

Cited by 1 publication

(1 citation statement)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Historical electricity consumption, socio-economic factors and the weather are taken as inputs in another work for the modelling of the electricity demand of Türkiye for the period of 2012-2016 using multilayer perceptron neural networks and it is concluded that the mean absolute percentage error of less than 2% is reached (Aydin and Toros, 2018). Similarly, past values of the demand are used as inputs in another study for the estimation of the hourly electricity demand of Türkiye for the period of 2018-2019 utilizing multilayer perceptron based neural networks where it is shown that high accuracy is achieved (Arslan et al, 2018). Historical temperature and consumption are used as inputs in another work for the multilayer perceptron and nonlinear artificial neural network based modelling of the electricity demand of an industrial region of Türkiye for the period of 2014-2016 where it is concluded that the multilayer perceptron based neural network performs better (Ozden and Ozturk, 2018).…”

Section: Literature Analysismentioning

confidence: 99%

Estimation of the Electricity Supply and Demand of Türkiye Employing Deep Learning Networks

Tuncsiper¹

2023

IJMRA

View full text Add to dashboard Cite

Electrical energy, one of the indispensable elements of the economic and social life, is an energy source that has to be provided without interruption in order to ensure the continuity of development. In this work, the monthly electricity supply and demand of Türkiye for the period of 2016-2022 are modelled using advanced machine learning techniques. An autoregressive deep learning network consisting of four hidden layers is developed for the modelling of the electricity supply and demand da ta having the form of multilayer perceptron artificial neural network. The electricity supply and demand data of Türkiye is gathered from the official sources and then investigated utilizing the seasonality analysis. Then, the deep learning network is developed in Python programming language. The 70% of the available electricity supply and demand data are used as the training data whereas the remaining 30% of the data is the test data. The randomized selection of the train and test data are performed using the classes available in the SciKit-Learn library of the Python programming language. The performance of the developed deep learning network model is evaluated both by plotting the actual and the modelled data on the same axis pair and by using performance metrics of the coefficient of determination, mean absolute error, mean absolute percentage error and the root mean square error. The mean absolute percentage errors of the models have the values of 2.29% and 2.17% for the electricity supply and demand data, respectively, indicating the success of the developed deep learning model for the modelling and the estimation of the electricity supply and demand data. The developed model is considered to be useful for the economic and technical energy planners.

show abstract

Section: Literature Analysismentioning

confidence: 99%

Estimation of the Electricity Supply and Demand of Türkiye Employing Deep Learning Networks

Tuncsiper¹

2023

IJMRA

View full text Add to dashboard Cite

show abstract

Short-Term Electricity Consumption Forecast Using Datasets of Various Granularities

Cited by 1 publication

References 14 publications

Estimation of the Electricity Supply and Demand of Türkiye Employing Deep Learning Networks

Estimation of the Electricity Supply and Demand of Türkiye Employing Deep Learning Networks

Contact Info

Product

Resources

About