Time-Stratified Analysis of Electricity Consumption: A Regression and
Neural Network Approach in the Context of Turkey
SİMGE YİĞİT1, SAFİYE TURGAY1, ÇİĞDEM CEBECİ2, ESMA SEDEF KARA3
1Department of Industrial Engineering,
Sakarya University,
54187, Esentepe Campus Serdivan-Sakarya,
TURKEY
2Department of Electrical Machinery and Material Supply,
Energy Branch Directorate,
Sakarya Municipality, Saski General Directorate,
Sakarya,
TURKEY
3Rüstempaşa Mahallesi,
İpekyolu Caddesi, No.120,
54600, Sapanca-Sakarya,
TURKEY
Abstract: - This study aims to apply seasonality and temporal effects in the analysis of electricity consumption
in Turkey as a case mixed with regression and neural network methodologies. The study goal is to increase
knowledge about the features and trending forces behind electricity usage which provide informed
recommendations for smart energy planning and regulation. Comparing and contrasting the regression and
neural network models makes it possible to carry out a thorough analysis of the merits and demerits of each
model. Moreover, the examination of the limits of the models and their performance in forecasting electricity
consumption patterns over the long term is done. The results of this study have a significant impact on power
forecasting techniques, and they have meaningful effects on the policymakers, planners and utilities in
Turkey. Understanding the story of the use of electricity around the world is very important for the
development of sustainable energy policies, resource provision, and the maintenance of reliable and smart
energy networks in the country.
Key-Words: - Electricity Consumption, Time-Stratified Analysis, Regression Modeling, Neural Network
Approach, Energy Forecasting, Turkey, Sustainable Energy Policies, Resource Optimization.
Received: March 17, 2023. Revised: January 2, 2024. Accepted: February 15, 2024. Published: April 2, 2024.
1 Introduction
The growing need for electricity in Turkey, being
connected with the mounting challenges of modern
electricity systems, requires a deep investigation
into consumers' behavior to welfare the formulation
of a desirable electricity plan and the enactment of
power policies. It is important to pay attention to
that as time-stratified analysis of electricity
consumption in the Turkey case is done with
traditional regression models and with the advanced
neural network methods. We hope to track the time-
based fluctuation of electricity demand in the hope
of exposing the driving factors of variation in usage
upshot.
Turkey's own geographical and climatic
peculiarities as well as her largest share of activities,
which determine electricity, demand structurally, all
are responsible for the development of a distinctive
electricity consumption profile. Knowing the
alterations in temporal attributes for instance
seasonality, temperature, and time-of-day is very
important for having optimal energy infrastructures,
and for making sure grid reliability and formulating
sustainable policies.
Then, as a solution to the determined
restrictions of traditional methods, the artificial
neural networks, which are the tools for non-linear
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
96
Volume 19, 2024
pattern building and complex relationship traffic,
are utilized.
Through applying a time-based model, the
research is committed to observing the complexity
of electricity consumption at different intervals,
seasonal factors, economic cycles, and possible
emotional issues. Net electricity generation is
defined as the total of all electricity production
minus all the same amount of electricity consumed
in a given region or country. The net electricity
consumption is significant to consider in terms of
energy demand, ensure the electricity demand and
supply balance. For operational purposes, net
electricity consumption is commonly measured in
megawatts or gigawatts (MWh/GWh). This gauge
indicates the difference between the total power
output formed over some time (for example, every
hour, day, month, or year) and the total power
consumption. The demand for grid electricity is
significant data for planners of energy, investments
in the area of infrastructure, and the design of
energy policy, just to mention a few. The net
electric power consumption of a portion or country
is used to identify the variations of its energy
demand, plan energy outputs, and utilize energy
resources properly. Data on net electricity
consumption was obtained from a variety of
sources, including energy companies, government
agencies, and international energy organizations.
This information is used in many areas such as
electricity consumption analysis, energy demand
forecasts, and energy policy development. However,
data on net electricity consumption is often
published with a lag and it is important to follow
official energy market data sources or producers for
up-to-date information.
The primary objective of this research is to
enhance the accuracy and depth of our
understanding of electricity consumption patterns in
Turkey through a two-fold approach: firstly,
employing a traditional regression model that
encompasses fundamental temporal variables, and
secondly, leveraging the power of neural networks
to capture complex, non-linear relationships
inherent in the data.
The first phase of our analysis involves
constructing a comprehensive regression model.
This model considers variables such as time of day,
day of the week, and seasonality, aiming to quantify
the impact of these factors on electricity
consumption. From historical data, the regression
model gives an insight into the time patterns of
consumption, which provides a background analysis
for later methods of more intricate mechanisms.
In the next phase, we move into the dimension
of neural networks, taking advantage of their
competence in unmasking perplexing patterns and
associations within huge datasets. The neural
network model adapts and learns from historical
patterns of consumption, which gives it high
accuracy for making predictions and discovering the
nuances that traditional regression models never see.
The main variables reflecting a study are; Real
Gross Domestic Product; Population; Quantity of
Vehicles; Foreign Trade; and Industry. Electricity
demand is a complex and multi-influencing
phenomenon, which depends on economic
activities, population growth, the effect of weather
elements, and technological developments. When
comprehending the volatile nature of electricity
usage, it is imperative for good energy planning and
sustainable development.
This article is about showing the temporal
variations of Turkey's electricity consumption using
time-stratified analysis, which is a combination of
traditional regression analysis as well as the
utilization of modern artificial neural networks
(ANNs).
The layout of the paper comes down to the
elements of the methodology, which in turn are
regression analysis and neural network
modeling. This time-layered analysis provides the
findings that are accommodative to the design and
formation of plans and strategies concerning energy,
which again, helps in policy formulation. Factors
that are seen to be substantial in one time. Period
could change in later times and this is where
regression analysis comes to the fore. On the other
hand, neural networks enable the researcher to have
a deep understanding of non-linear relationships
between factors. Through this, predictions that are
more accurate are achieved. The anticipated effects
of these findings on energy policy or planning
include the discussion of effective approaches and
the prospects for more competent forecasting and
sound decision-making in the volatile reality of
consumption are illustrated. This study adds to the
emerging branch in energy analytics of investigating
the performance of regression models as a simple
alternative to deep neural networks in electricity
consumption prediction. As such, they bring
significant impacts for the above-mentioned actors,
that is, energy planners, policymakers, and utility
companies, which in turn are used as important and
relevant inputs in allocating resources and
developing sustainable energy strategies in Turkey.
As we delve into the depths of time-stratified
electricity consumption analysis, the subsequent
sections of this research will unfold the
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
97
Volume 19, 2024
methodology, data, results, and discussions,
ultimately leading to a nuanced understanding of the
temporal intricacies shaping Turkey's electricity
landscape.
The remainder of this paper is structured as
follows: Section 2 includes an overview of the
pertinent literature. Section 3 defines the model
definition and formulation, encompassing the
mathematical model of the time-stratified analysis
of electricity consumption. Section 4 structures a
case study, summarizing the key results of our
proposed approach in comparison to the current
state of affairs. Finally, in Section 5, we present our
concluding remarks.
2 Literature Survey
The literature survey explores existing research
related to time-stratified analysis of electricity
consumption, with a focus on regression and neural
network approaches, within the specific context of
Turkey. Some of the studies cover the temporal
analysis of electricity consumption which
emphasizes the importance of considering temporal
dynamics for accurate forecasting and efficient
energy planning, [1], [2], [3], [4], [5]. Some of the
studies highlight the ability of neural networks to
capture non-linear relationships and intricate
patterns, contributing to more accurate predictions
in electricity consumption modeling, [6], [7], [8],
[9], [10]. Some of the researchers have explored the
influence of geographical and socioeconomic
factors on electricity consumption, [11], [12]. Given
Turkey's unique characteristics, understanding how
these factors interact with temporal dynamics
becomes crucial for tailored energy policies. Some
of these works shed light on the country's energy
landscape, providing valuable insights into
consumption patterns and trends that inform the
present research, [13], [14], [15]. Comparative
analyses between regression models and neural
networks in the context of electricity consumption
are scarce, [16], [17], [18], [19]. The authors
discussing the implications of electricity
consumption patterns on energy policy offer a
broader perspective, [20], [21], [22]. The present
research aims to contribute to this discourse by
providing insights specifically tailored to the
Turkish context. Advancements in time series
analysis have been crucial for refining
methodologies in electricity consumption studies.
The integration of advanced artificial intelligence in
the current technology is an effort made to improve
the accuracy of timeline forecasts and deepen
understanding of it, [23], [24], [25]. The other
investigations address the place of temperature and
or weather in total energy consumption. In the spirit
of the diversity of weather in Turkey, these
correlations are of significant value to the accuracy
of the modeling and forecasting, [26], [27], [28],
[29], [30].
It is with this basic objective that the current
research will bring contributions by building upon
existing knowledge and making a definitional
addition to Electricity Consumption Analysis, which
is valid in the Turkish context.
3 Methodology
Herewith the given analysis is undertaken which
focuses on electricity consumption in Turkey.
Grasping a holistic view by using both traditional
regression analysis and advanced artificial neural
networks (ANNs). The main purpose of the research
is to find hidden relationships and identify changes
that occur inside the time domain of electricity
demand. This kind of information is very valuable
as it tries to support the future policymakers, as well
as all energy planners, and of course their
stakeholders.
Regressions analysis is very stable. Through an
ordered review of historical data, key variables such
as consumption of energy with time can be easily
identified. Then, besides the stem learning
algorithms like regression and correlation analysis,
which are excellent in terms of capturing any non-
linearity, the artificial neural networks are applied to
improve the forecasting accuracy.
The research uses a data set that covers a particular
period, which is then divided further into numerous
periods to ensure seasonality, trend, and anomaly
injection of electricity consumption patterns being
captured in totality. Thus, by the integration of the
regression and neural network techniques, this
research seeks a detailed examination of time-
dependent features of Turkey's electricity
demand. The end purpose is to increase our
knowledge of how the different parameters
influence the electricity consumption in Turkey
during different periods. By using the strengths of
both interpretability associated with regression
analysis and the flexibility provided by neural
networks, this methodological dual approach allows
us to gain a deeper understanding of demand
response and how different factors affect it in the
temporal sense. Finally, the study is considered as a
useful tool for decision-makers in the energy sector,
and it will help prepare effective performance plans
and permanent actions that are used by energy
management planners in Turkey.
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
98
Volume 19, 2024
In electricity consumption time stratified analysis,
regression and neural network approaches on the
one hand provide a flexible and integrating
methodology. The following paragraph is about the
regression model, as well as the neural network.
3.1 Mathematical Model
3.1.1 Regression Model
The regression model represented as a linear
equation:
Y=0+1X1+2X2+…+nXn+
Where
Y is the dependent variable (electricity
consumption)
0 is the intercept.
1+2+…+n are the coefficients for the
independent variables
X1+X2+…+Xn, representing the time of day, day of
the week, etc.)
is the error term.
For example, considering time of day (TOD),
day of the week (DOW), and seasonal factors
(SEASON), the regression equation might be
Consumtion=0+TOD.TOD+DOW.DOW+SEASON.DOSEAS
ON+
3.1.2 Neural Network Model
Let's consider a simple feedforward neural network:
Where
is the predicted electricity consumption.
X is the input vector containing features such as
time of day, day of the week, etc.
f() is the activation function (e.g., ReLU for hidden
layers, linear for output layer).
and are weight matrices.
and are bias vectors.
For instance, with a single hidden layer, the
equation might look like:
3.1.3 Training
Both models involve training by minimizing a loss
function. For the regression model, it might be
Mean Squared Error (MSE):
For the neural network, it could be the same
MSE or another suitable loss function.
3.1.4 Optimization
Optimization algorithms, such as gradient descent,
are used to adjust the parameters (β for regression,
and b for neural network) to minimize the loss:
Interpretability
- Regression Model
o The coefficients (β) provide insights
into the impact of each temporal
variable on electricity consumption.
- Neural Network
o Techniques like SHAP values or
layer-wise relevance propagation
can be employed for
interpretability, attributing
predictions to input features.
In practice, the complexity of neural network
architectures can vary based on the problem's
intricacy. This simplified model representation
provides a foundation for understanding the core
mathematical concepts involved in both regression
and neural network approaches.
Policymakers can use temporal insights to
implement time-specific energy-saving measures,
incentivizing consumers to reduce consumption
during peak periods. The study's findings guide
infrastructure planning, helping to design systems
capable of handling peak demand periods
efficiently. Incorporating neural network predictions
into existing forecasting models can enhance the
accuracy of electricity consumption predictions,
supporting more effective energy planning.
The study's findings are contingent on the
quality and availability of historical data. Improved
data collection methods could enhance the accuracy
of predictions. While the neural network
demonstrated superior performance, its complexity
may pose challenges for interpretation. Future
research could focus on developing hybrid models
for improved interpretability. External factors such
as economic changes or policy shifts were not
explicitly considered. Future studies could explore
the integration of external variables for a more
holistic analysis.
4 Case Study
In this study, we conducted a time-stratified analysis
of electricity consumption in Turkey, employing
both regression and neural network approaches. The
research aimed to enhance our understanding of
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
99
Volume 19, 2024
temporal patterns and drivers behind electricity
consumption, providing valuable insights for energy
planning and policy formulation (Figure 1).
Fig. 1: The proposed model outline
The regression model revealed significant
relationships between temporal variables (time of
day, day of the week, and seasonality) and
electricity consumption. Interpretation of
coefficients highlighted the impact of specific time-
related factors on consumption patterns. The neural
network outperformed the regression model in
capturing complex, non-linear relationships within
the electricity consumption dataset. The model
demonstrated adaptability to intricate patterns,
providing more accurate predictions than the
traditional regression approach.
Comparative analysis highlighted the strengths
and limitations of each approach. The regression
model offered interpretability, while the neural
network excelled in capturing intricate patterns.
Both models demonstrated generalizability,
effectively predicting electricity consumption trends
in unseen data. The neural network, however,
showcased superior adaptability to diverse patterns.
Insights from this study have direct implications for
energy policy in Turkey. Understanding temporal
dynamics is crucial for optimizing resource
allocation, ensuring grid reliability, and developing
sustainable energy strategies. Turkey's electricity
consumption between 1975 and 2021 and the
independent variables [ (Population, Gross
Domestic Product, Number of Vehicles, Foreign
Trade ($), Industry (TL)] that are thought to affect
this consumption amount are used. The study
proceeds in line with two objectives. In the first
objective, the effects of the independent variables on
the net consumption amount are analyzed and it is
examined at how many times which variable affects
it. Secondly, it is aimed to determine whether these
methods are successful in predicting net electricity
consumption and to calculate which method is more
successful with performance measures after
prediction by regression analysis and artificial
neural networks method. Prediction values were
compared with actual values and error metrics were
calculated. The data were analyzed with Google
Colab.
By systematically implementing this
methodology, the study aims to provide a
comprehensive understanding of time-stratified
electricity consumption in Turkey, comparing the
efficacy of traditional regression models with
advanced neural network approaches.
The proposed algorithm presents a hybrid model
to capture the temporal dynamics of electricity
consumption in Turkey by combining regression
and neural network approaches. Through extensive
analysis and model comparison, the algorithm aims
to provide accurate forecasts and valuable insights
for sustainable energy planning and policy
formulation.
1. Obtaining and organizing data
Data on the amount of electricity consumption and
variables were found in bibliographies containing
and analyzing various statistical documentation.
These data are organized in an Excel file and
converted into tables (Figure 2).
Fig. 2: Data set sample
2. Analyzing the data
The data were analyzed through Google Colab file,
missing data detection, data categorization,
correlations, and their effects on consumption
amounts were analyzed (in Figure 3, Figure 4,
Figure 5, Figure 6, Figure 7).
3. Selection of the appropriate solution method
Since the problem is aimed at forecasting,
regression, and artificial neural network models
were selected among the models that could be
suitable for this.
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
100
Volume 19, 2024
Fig. 3: The relationship between Net Consumption
and Real GDP
Fig. 4: The relationship between Net Consumption
and Population
Fig. 5: The relationship between Net Consumption
and Vehicles
Fig. 6: The relationship between Net Consumption
and Foreign Trade
Fig. 7: The Relationship between Net Consumption
and Industry
4. Regression and ANN modeling
The problem was modeled using the necessary
coding on Google Colab (Table 1).
Table 1. Prediction results and regression results
5. Evaluation of performances
For the performance values of the models, their
scores were analyzed and the mean squared error
and the margin of error and the differences between
them were found in Figure 8 and Table 2.
Fig. 8: Statistical Results
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
101
Volume 19, 2024
Table 2. Performance analysis results
As a result, the regression model was found to be
more successful.
5 Conclusion
In this study, we conducted a time-stratified analysis
of electricity consumption in Turkey using both
regression analysis and artificial neural network
(ANN) models. In this context, the goal was to show
the different trends in electricity consumption and
think of possible power planning strategies and
ideas for new energy policies. Coming together
prediction techniques by using regression and neural
network approaches, a model hybrid is improved for
achieving more effectiveness in forecasting and
capturing electricity consumption dynamics
complexity.
In the case of the Mean Squared Error, it
concluded that the model of regression was ideal for
this set of data. The network of Artificial Neural
Network (ANN) is designed in such a way to be
able to analyze the most complicated data sets that
are not analyzed by regression analysis.
In the analysis, it very complicated periodical flow
of power use that had changing nature of the time
and difference of the periods, respectively.
Seasonal phenomena, economic difficulties as
well as activity during a specific period had quite a
special influence on the electricity
demand. Regression Models presented various
factors concerning electricity consumption and the
economic impacts on selected
determinants. However, economic indicators,
population growth, and climate factors were at
different degrees of significance in regard to
consumption changes at various times. The
networks of neurons with their abilities to show
non-linear correlations were telling us something
hidden among the complexity of data. They
practiced their mastery of not only technical but also
macroeconomic feedback, especially throughout
periods of non-linear growth in consumption
rates. Sometimes, combining the prognostic strength
of regression and neural network results in an even
more accurate and better-performing performing the
hybrid-forecasting model. The coefficient value (α)
optimized for harmony to take from both
procedures, thus, reducing prediction error and
making the model more comprehensive.
The result obtained from this analysis just can
contribute to a more precise forecasting of the
amount of electricity, that people consume in
Turkey. Thus, energy turbines and officials can
make use of these forecasts to plan well resources
based on the use and development of physical
infrastructures. Knowledge of seasonality in
electricity demand enables purposeful strategy in a
place where high seasonal demand is addressed and
conserving off-seasons to make. The time-step
modeling method allows for identifying and
anticipating the incidence of irregularities in
consumption, which is crucial for increasing the
adaptability and resilience of the energy system.
In summary, this study reveals informative
knowledge about the regional, seasonal, and
organizational factors involving electricity
consumption in Turkey.
The set of regression analysis, neural network
modeling, and hybrid approaches is not only an
effective model but also a very robust tool that is
well-suited for understanding and forecasting the
dynamic behavior of energy demand over time. This
is linked as well with the inspiration of policies that
recognize the use of renewable energies in Turkey.
The benefit will ultimately be a sustainable and
resilient energy future for the country.
Acknowledgement:
It is an optional section where the authors may write
a short text on what should be acknowledged
regarding their manuscript.
References:
[1] Nsangou, J.C., Kenfack, J., Nzotcha, U.,
Ekam, P.S.N., Voufo, J., Tamo, T.T.,
Explaining household electricity consumption
using quantile regression, decision tree and
artificial neural network, Energy, Vol. 250, 1
July 2022, 123856.
[2] Schneider, N., Strielkowski, W., Modelling
the unit root properties of electricity data—A
general note on time-domain applications,
Physica A: Statistical Mechanics and its
Applications, Vol. 618, 15 May 2023,
128685.
[3] Sharafi, S., Kazemi, A., Amiri, Z., Estimating
energy consumption and GHG emissions in
crop production: A machine learning
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
102
Volume 19, 2024
approach, Journal of Cleaner Production, Vol.
408, 1 July 2023, 137242.
[4] Bacanin, N., Jovanovic, L., Zivkovic, M.,
Kandasamy, V., Antonijevic, M., Deveci, M.,
Strumberger, I., Multivariate energy
forecasting via metaheuristic tuned long-short
term memory and gated recurrent unit neural
networks, Information Sciences, Vol. 642,
September 2023, 119122.
[5] Kesriklioğlu, E., Oktay, E., Karaaslan, A.,
Predicting total household energy
expenditures using ensemble learning
methods, Energy, Vol. 276, 1 August 2023,
127581.
[6] Michael P. B. , Abrasaldo, S. J. Z., Andreas
W. Kempa-Liehr, A.W., A systematic review
of data analytics applications in above-ground
geothermal energy operations, Renewable and
Sustainable Energy Reviews, Vol. 189, Part
B, January 2024, 113998.
[7] Mehmood, U. M., Chun, D., Shan, Z.,
Hyunjoo Han, H., Jeon, G., Chen, K., A
review of the applications of artificial
intelligence and big data to buildings for
energy-efficiency and a comfortable indoor
living environment, Energy and Buildings,
Vol. 202, 1 November 2019, 109383.
[8] Tso, G.K.F., Yau, K.K.W., Predicting
electricity energy consumption: A comparison
of regression analysis, decision tree and
neural networks, Energy, Vol. 32, Issue
9, September 2007, Pages 1761-1768.
[9] Oreshkin, B.N., Dudek, G., Pełka, P., Turkina,
R., N-BEATS neural network for mid-term
electricity load forecasting, Applied Energy,
Vol. 293, 1 July 2021, 116918.
[10] Civak, H., Küren, C., Turgay, S., Examining
the effects of COVID-19 Data with Panel
Data Analysis, Social Medicine and Health
Management (2021) Vol. 2: 1-16 Clausius
Scientific Press, Canada, DOI:
10.23977/socmhm.2021.020101 ISSN 2616-
2210.
[11] Kheiri, F., A review on optimization methods
applied in energy-efficient building geometry
and envelope design, Renewable and
Sustainable Energy Reviews, Vol.
92, September 2018, Pages 897-920.
[12] Shine, P., Murphy, M.D., Upton, J., Scully,
T., Machine-learning algorithms for
predicting on-farm direct water and electricity
consumption on pasture based dairy farms,
Computers and Electronics in Agriculture,
Vol. 150, July 2018, Pages 74-87.
[13] Marriette Sakah, M., Can, S.R., Diawuo, F.A.,
Sedzro, M.D., Kuhn, C., A study of appliance
ownership and electricity consumption
determinants in urban Ghanaian households,
Sustainable Cities and Society, Vol.
44, January 2019, Pages 559-581.
[14] Saryazdi, S.M.E., Etemad, A., Shafaat, A.,
Bahman, A.M., A comprehensive review and
sensitivity analysis of the factors affecting the
performance of buildings equipped with
Variable Refrigerant Flow system in Middle
East climates, Renewable and Sustainable
Energy Reviews, Vol. 191, March 2024,
114131.
[15] Zuin, G., Buechler, R., Sun, T., Zanocco, C.,
Galuppo, F., Veloso, A., Rajagopal, R.,
Extreme event counterfactual analysis of
electricity consumption in Brazil: Historical
impacts and future outlook under climate
change, Energy, Vol. 281, 15 October 2023,
128101.
[16] Lazzari, F., Mor, G., Cipriano, J., Gabaldon,
E., Grillone, B., Chemisana, D., Solsona, F.,
User behaviour models to forecast electricity
consumption of residential customers based
on smart metering data, Energy Reports, Vol.
8, November 2022, Pages 3680-3691.
[17] Lahmar, S., Maalmi, M., Idchabani, R.,
Investigating adaptive sampling strategies for
optimal building energy performance using
artificial neural networks and kriging
surrogate models, Journal of Building
Engineering, Vol. 82, 1 April 2024, 108341
[18] Moustafa, M., Ruifeng, T., Wen, J., Bo, W.,
Ullah, A., Mohamad, H.A.E., Cheng, H.,
Modeling of wavy water film by application
of artificial neural network - a state of art
study, Nuclear Engineering and Design, Vol.
417, February 2024, 112731.
[19] Zou, Y., Lin, Z., Li, D., Liu, Z.C.,
Advancements in Artificial Neural Networks
for health management of energy storage
lithium-ion batteries: A comprehensive
review, Journal of Energy Storage, Vol. 73,
Part C, 15 December 2023, 109069.
[20] Palaniappan, S., Karuppannan, S., Velusamy,
D., Categorization of Indian residential
consumers electrical energy consumption
pattern using clustering and classification
techniques, Energy, Vol. 289, 15 February
2024, 129992.
[21] Lee, D., Ooka, R., Ikeda, S., Choi, W., Kwak,
Y., Model predictive control of building
energy systems with thermal energy storage in
response to occupancy variations and time-
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
103
Volume 19, 2024
variant electricity prices, Energy and
Buildings, Vol. 225, 15 October 2020,
110291.
[22] Mohapatra, S.K., Mishra,S., Tripathy, H.K.,
Alkhayyat, A., A sustainable data-driven
energy consumption assessment model for
building infrastructures in resource constraint
environment, Sustainable Energy
Technologies and Assessments, Vol. 53, Part
C, October 2022, 102697
[23] Ghadami, N., Gheibi, M., Kian, Z., Faramarz,
M.G., Naghedi, R., Eftekhari, M., Fathollahi-
Fard, A.M., Dulebenets, M.A., Tian, G.,
Implementation of solar energy in smart cities
using an integration of artificial neural
network, photovoltaic system and classical
Delphi methods, Sustainable Cities and
Society, Vol. 74, November 2021, 103149
[24] Sefeedpari, P., Rafiee, S., Akram, A., Chau,
K., Pishgar-Komleh, S.H., Prophesying egg
production based on energy consumption
using multi-layered adaptive neural fuzzy
inference system approach, Computers and
Electronics in Agriculture, Vol.
131, December 2016, Pages 10-19
[25] Kayali, S., Turgay, S., Predictive Analytics
for Stock and Demand Balance Using Deep
Q-Learning Algorithm. Data and Knowledge
Engineering (2023) Vol. 1: 1-10. DOI:
http://dx.doi.org/10.23977/datake.2023.01010
1.
[26] Ren, S., Hu, W., Bradbury, K., Harrison-
Atlas, D., Valeri, L.M., Murray, B., Malof,
J.M., Automated Extraction of Energy
Systems Information from Remotely Sensed
Data: A Review and Analysis, Applied
Energy, Vol. 326, 15 November 2022, 119876
[27] Blaga, R., Sabadus, A., Stefu, N., Dughir, C.,
Paulescu, M., Badescu, V., A current
perspective on the accuracy of incoming solar
energy forecasting, Progress in Energy and
Combustion Science, Vol. 70, January 2019,
Pages 119-144
[28] Hong, Y.Y., Paulo C.L., Rioflorido, P.,
Zhang, W., Hybrid deep learning and
quantum-inspired neural network for day-
ahead spatiotemporal wind speed forecasting,
Expert Systems with Applications, Vol. 241, 1
May 2024, 122645
[29] Taşkın, H., Kubat, C., Topal, B., Turgay, S.,
Comparison Between OR/Opt Techniques
and Int. Methods in Manufacturing Systems
Modelling with Fuzzy Logic International
Journal of Intelligent Manufacturing, 15, 517-
526 (2004).
[30] Nsangou, J.C.., Kenfack, J., Nzotcha, U.,
Ekam, P.S:N., Voufo, J., Tamo, T.T.,
Explaining household electricity consumption
using quantile regression, decision tree and
artificial neural network, Energy, Vol.
250,2022,123856.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
The authors equally contributed to the present
research, at all stages from the formulation of the
problem to the final findings and solution.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflicts of interest to declare.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
WSEAS TRANSACTIONS on POWER SYSTEMS
DOI: 10.37394/232016.2024.19.12
Si
mge Yi
ği
t, Safi
ye Turgay,
Çi
ğdem Cebeci
, Esma Sedef Kara
E-ISSN: 2224-350X
104
Volume 19, 2024
