applied

sciences
Article
Performance Evaluation of Two Machine Learning
Techniques in Heating and Cooling Loads Forecasting
of Residential Buildings
Arash Moradzadeh 1 , Amin Mansour-Saatloo 1 , Behnam Mohammadi-Ivatloo 1,2, *
and Amjad Anvari-Moghaddam 1,3, *
1 Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz 5166616471, Iran;
arash.moradzadeh@tabrizu.ac.ir (A.M.); amin_mnsr97@ms.tabrizu.ac.ir (A.M.-S.)
2 Institute of Research and Development, Duy Tan University, Da Nang 550000, Vietnam
3 Department of Energy Technology, Aalborg University, 9220 Aalborg, Denmark
* Correspondence: bmohammadi@tabrizu.ac.ir (B.M.-I.); aam@et.aau.dk (A.A.-M.)




Received: 29 April 2020; Accepted: 29 May 2020; Published: 31 May 2020


Abstract: Nowadays, since energy management of buildings contributes to the operation cost, many
efforts are made to optimize the energy consumption of buildings. In addition, the most consumed
energy in the buildings is assigned to the indoor heating and cooling comforts. In this regard,
this paper proposes a heating and cooling load forecasting methodology, which by taking this
methodology into the account energy consumption of the buildings can be optimized. Multilayer
perceptron (MLP) and support vector regression (SVR) for the heating and cooling load forecasting of
residential buildings are employed. MLP and SVR are the applications of artificial neural networks
and machine learning, respectively. These methods commonly are used for modeling and regression
and produce a linear mapping between input and output variables. Proposed methods are taught
using training data pertaining to the characteristics of each sample in the dataset. To apply the
proposed methods, a simulated dataset will be used, in which the technical parameters of the
building are used as input variables and heating and cooling loads are selected as output variables
for each network. Finally, the simulation and numerical results illustrates the effectiveness of the
proposed methodologies.

Keywords: energy management; load forecasting; heating and cooling; machine learning; multi-layer
perceptron (MLP); support vector regression (SVR)

1. Introduction
Increasing the number of cities and their populations throughout the world requires a great
deal of energy to meet the needs of citizens. Recent studies have predicted that the population of
cities will increase up to five billion by 2030 [1]. Nearly 40% of total energy consumption is related
to the dwellings, and other building types constitute just a fraction of the buildings [2]. Supplying
energy to the citizens requires associated resources in which limited sources are available. As the
consumption of residential buildings forms a great amount of demand, regarding social welfare,
residential consumption should be monitored and controlled [3]. On the other hand, heating and
cooling are the most crucial energy sources among citizens, so in this regard usage of these energies
should be managed.
Managing and optimizing the energy consumption of buildings requires having complete
information about the performance of the building and environmental factors. Electricity, gas and
heating supply are the most important resources of energy in a building, but the important final use

Appl. Sci. 2020, 10, 3829; doi:10.3390/app10113829 www.mdpi.com/journal/applsci

Appl. Sci. 2020, 10, 3829 2 of 12

applications are elevators, heating ventilation and air conditioning (HVAC), domestic hot water, and
so on. Among the aforementioned energy sources, optimal operation of HVAC and indoor condition
supply are two important factors in evaluating building energy performance [4,5]. HVAC, as a basic
infrastructure in the building, plays an important role by changing the amount of internal cooling and
heating loads of residential buildings. Despite the need for this system in buildings, there is a major
concern that about 40% of all energy, especially in office buildings, is consumed by this system [6,7].
Forecasting the thermal loads plays an important role in optimizing the cooling and heating cost
of the buildings, as the deviation from the optimally scheduled values will increase the total cost
considerably [8].
Energy forecasting is a way to reach optimal operation of HVAC and energy management of
residential buildings. In this regard, residential buildings’ consumption patterns could be predicted [9].
Nowadays, by developing technology many small-scale smart devices and building management systems
(BMSs) can be installed on the residential buildings sites in order to monitor and record the load patterns
of the buildings, and also environmental characteristics that have a high impact on the energy forecasting.
Using such data, building consumption patterns could be predicted and controlled hourly. In addition,
the necessity of energy forecasting has been expressed by some researches e.g., a review study of thermal
energy consumption in the buildings pertinent to the comfort was introduced in [10] and the purpose of
the study was showing how social-economic, fuel mix and climate change are affected by thermal energy
comfort. Moreover, by load forecasting, buildings can schedule for the next day, not only to participate in
the demand–response programs [11–13] but also to participate in energy trading [14,15].
So far, many studies have been done in order to evaluate load forecasting of the buildings. In a
valuable study [16], an integrated design approach has been utilized to estimate life-cycle energy
savings, cost-effectiveness of energy efficiency measures in new buildings, and carbon emission
reduction. In [17], a multi-objective optimization for energy refurbishments of existing buildings
through energy efficiency measures and HVAC systems have been carried out using a genetic algorithm
coupled with a dynamic simulation tool. Predictor methods for heating load based on artificial neural
networks (ANN) have been evaluated in [18] for office buildings where the impact of data size and
dimensionality in ANN was inspected. In order for heating, ventilation, and air-conditioning (HVAC)
system optimization in [19], electricity load forecasting based on ANN has been studied. Among
three utilized algorithms such as Levenberg-Marquardt, Scaled Conjugate gradient back-propagation,
and Bayesian Regularization (BR), the BR-based ANN showed the best performance. Another study
proposed the energy forecasting method using statistical analysis for heating and cooling of an
office building [20]. In [21], four hybrid techniques based on artificial neural network (ANN) and
meta-heuristic algorithms such as artificial bee colony (ABC) optimization, particle swarm optimization
(PSO), imperialist competitive algorithm (ICA), and genetic algorithm (GA) have been suggested for
forecasting the heating load of buildings’ energy efficiency. Forecasting the cooling load has been
done in [22] using a probabilistic entropy-based neural (PENN) method. Short-term cooling load
prediction in order to optimize the operation of HVAC systems and energy efficiency measures in
buildings has been done in [23] using multiple nonlinear regression (MNR), auto regressive (AR),
and autoregressive with exogenous (ARX) models. In [24], the thermal comfort reduction of the energy
consumption in the building by 36.5% was performed via a feedforward neural network (FFNN). A
decision tree method has been suggested in [25] for energy demand forecasting and energy efficiency
measures of a residential building. A comparative study of forecasting methods for heating and
cooling load was done in [26], where machine learning techniques such as a deep neural network
(DNN), gradient boosted machine (GBM), Gaussian process regression (GPR) and minimax probability
machine regression (MPMR) were compared with each other. In [27], prediction of the cooling and
heating loads of the building were done using ANN, classification and regression tree (CART), general
linear regression (GLR), and chi-squared automatic interaction detector (CHAID). In the same work,
the technical characteristics of the building were considered as input to the networks. In [28], sixteen
residential buildings were evaluated in terms of heating and cooling energy consumption forecasting
Appl. Sci. 2020, 10, 3829 3 of 12

via adaptive linear time-series models. Likewise, cooling load forecasting based on data mining
techniques was proposed in [29] to help design a more efficient building management system (BMS).
In [30], the BMS based on electrical, economic, and ecological optimization using a genetic algorithm
was introduced to improve energy efficiency of the buildings. General regression neural network
(GRNN) has also been employed in [31] for cooling energy forecasting to optimize HVAC heat storage
of public buildings.
In most of the aforementioned research works, meteorological data was used as an indicator
and input for forecasting the cooling and heating loads of residential buildings. It is undeniable that
environmental and meteorological factors do not affect the cooling and heating loads of residential
buildings, but sometimes abrupt changes in weather could disrupt energy forecasting equations,
reducing the accuracy coefficient and increasing the error in the energy forecasting operation. In this
paper, high-precision prediction of cooling and heating loads of a building was done by using
multilayer perceptron (MLP) and support vector regression (SVR) methods. A set of data on structural
characteristics of the building was considered as an input variable, while the amount of cooling and
heating load was considered as an output variable. Using this data and creating a linear mapping
between input and output variables via the proposed methods, it is possible to make a more accurate
prediction of cooling and heating loads.
The rest of the paper is organized as follows: Section 2 describes the case study and dataset.
Section 3 represents the employed methodologies and techniques. Section 4 includes the simulation
and numerical results and finally, Section 5 concludes the paper.

2. Case Study
The dataset used in this work was created by Tsanas and Xifara [32]. Twelve different buildings
were simulated in Ecotect software to generate the dataset. The glazing area, distribution of the glazing
area and the orientation are the parameters that make the buildings different from each other. Each
building was simulated using eighteen preliminary cubes (3.5 × 3.5 × 3.5 m3 ) with the same materials
for all buildings. The newest and most common materials in the building construction industry were
selected for each of the eighteen elements so that the materials used for each of these elements were the
same for all forms of construction. In the design process, three types of glazing areas such as 10%, 25%,
and 40% were used as percentages of the floor area. In addition, it was assumed that buildings were
in Greece, Athens. Sixty percent humidity, 0.3 m/s wind speed, lightning level of 300 1x and 0.6 clo
of clothing were considered as internal design conditions during simulation, while the infiltration
rate was set to 0.5 for the air change rate with a wind sensitivity of 0.25 air changer per hour. The
dataset includes 768 samples with eight features for each sample, namely x1 , x2 , . . . , x8 and y1 , y2 as
decision variables, which are listed in Table 1 [21,32]. This work aims to forecast y1 as the heating
load and y2 as the cooling load using the aforementioned features as decision variables. Although the
dataset was generated via simulation, it is notable that the proposed methods are applicable to the
real-world dataset.

Table 1. Details of input and output data.

Mathematical Symbol Variables

x1 Relative compactness
x2 Surface area
x3 Wall area
x4 Roof area
x5 Overall height
x6 Orientation
x7 Glazing area
x8 Glazing area distribution
y1 Heating load
y2 Cooling load
 y2 Cooling load

3. Methods
Artificial
Appl. Sci. neural network (ANN) and machine learning algorithms as powerful tools in data4 min‐
2020, 10, 3829 of 12
ing were employed to do the modelling and forecasting tasks [33], [34]. In this work, MLP and SVR
were used as two application models of these algorithms to create a linear mapping between the
3. Methods
technical parameters of building and the cooling and heating loads of the building in order to forecast
Artificial neural
the load/energy. network
In the (ANN)
following, eachand machine
of the learning
proposed methodsalgorithms
are brieflyas introduced.
powerful tools in data
mining were employed to do the modelling and forecasting tasks [33,34]. In this work, MLP and
3.1. Multilayer
SVR were used perceptron (MLP) models of these algorithms to create a linear mapping between the
as two application
technical
MLP parameters
has a fullyofconnected
building and
layerthe cooling i.e.,
structure, and each
heating loadsinofa the
neuron building
layer in order
is connected to forecast
to all neurons
the load/energy. In the following, each of the proposed methods are briefly introduced.
in the next and previous layers. The schematic of the MLP structure is shown in Figure 1, which
illustrates
3.1. a nonlinear
Multilayer mapping
Perceptron (MLP) between the input vector and the output vector [35]. The neurons are
connected through weights, and output signals are generated by a nonlinear transfer function [36].
MLP has a fully connected layer structure, i.e., each neuron in a layer is connected to all neurons
N
Y  f (b  w l x l )
illustrates a nonlinear mapping between the input vectorl 1

in the next and previous layers. The schematic of the MLP structure is shown in Figure 1, which
(1)
and the output vector [35]. The neurons are
connected through weights, and output signals are generated by a nonlinear transfer function [36].
In equation (1) Y and x are the output and input signals, respectively, f is the nonlinear trans‐
fer function, b and w are the bias and weight vectors,XN respectively, and N is the total number of the
Y =through
inputs. Since MLP has the ability to learn f (b + training,
wl xl ) a dataset with known input vector and
(1)
l=1
output vector is required in which the weight vector is adjusted according to the output signals
through training [37].

.
Figure 1. Basic structure of multilayer perceptron (MLP).
Figure 1. Basic structure of multilayer perceptron (MLP).
In Equation (1) Y and x are the output and input signals, respectively, f is the nonlinear transfer
3.2. Support vector regression (SVR)
function, b and w are the bias and weight vectors, respectively, and N is the total number of the inputs.
Since SVR
MLPishasonethe
of ability to learn
the training toolsthrough
whichtraining, a dataset
was developed with
from theknown
support input vector
vector and output
machine (SVM).
vector is required
The principle in which
of SVR the weight
is depicted in Figure vector 2. Inisthis
adjusted according
work, ε‐SVR to the output
is employed for thesignals
trainingthrough
of data.
training
ε‐SVR is[37].
a classic model of SVR with the aim of finding a flat function, which has a small (ε) error
from the obtained target [38]. In the case of SVR, the following function is trained using given training
3.2. Support Vector Regression (SVR)
data such as {(x 1 , y 1 ), (x 2 , y 2 ),...,(x l , y l ),...(x N , y N )}    , l  {1, 2,..., N } , where  illus‐
SVR
trates theisspace
one of
ofthe
thetraining tools which was developed from the support vector machine (SVM).
input patterns:
The principle of SVR is depicted in Figure 2. In this work, ε-SVR is employed for the training of data.
ε-SVR is a classic model of SVR with f (the  wof,finding
x )aim x  b ;w  function,
a flat , b  which has a small (ε) error from (2)
the obtained target [38]. In the case of SVR, the following function is trained using given training
data such as (x1 , y1 ), (x2 , y2 ), . . . , (xl , yl ), . . . (xN , yN ) ⊂ χ × R, l = {1, 2, . . . , N}, where χ illustrates the


space of the input patterns:

f (x) = hw, xi + b; w ∈ χ, b ∈ R (2)
Appl. Sci. 2020, 10, 3829 5 of 12

where b is bias, and hw, xi represents the linear function of fitting input space to the feature space.
Equation (3) is using to minimizing the risk function as follows:

N
1 X
R= kwk2 + c ( y1 , hw, xi) (3)
2
i=1
Appl. Sci. 2020, 10, 3829 6 of 13

Figure Theprinciple
2. The
Figure 2. principleofofsupport
supportvector
vector regression
regression (SVR).
(SVR).

In Equation
4. Simulation the selected loss function and c > 0 specifies the tradeoff between the smoothness
(3),results
and
of f and allowed deviation larger than ε. In order to deal with the problem, the minimization problem
The MLP and SVR networks are designed to predict the cooling and heating load. Each of these
of (4) must be solved.
networks was trained using a dataset as input. In thisX Npaper, 85% (658 samples) of the data were used
1 2 the remaining 15%
to train and validate the proposed methods and
min kwk + c (ξ + ξ∗l ) (110 samples) were used for test‐ (4)
ing. In the first stage, each network required 2 a preliminaryl design to determine the number of neu‐
l=1
rons in the hidden layer and the coefficients  of the network. After designing each network, the
 yl − hw, xl i − b ≤ ε + ξl

amount of training and testing data for  the network was determined. In this work, 70% of the samples
≤ ε + ξ∗l the training phase of each network.

 hw,
s.t.30% li +
xtest b − to
yl validate

were considered as training data and  as data
ξl regression
+ ξ∗l ≥ 0 algorithm, the results need to be evalu‐


After training and testing each neural network or

ated. To
where, ξl , do this,the
ξ∗l are correlation coefficient
slack variables which(R), tackle
mean squared error (MSE),
the infeasible root mean
constraints. squared
In order error the
to solve
(RMSE), and problem,
optimization mean absolute errorproblem
the dual (MAE) can be used
of the as statistical
(4) can be derived performance metrics. Each
using the Lagrange of the In
function.
mentioned indices are calculated according to the following equations [42].
addition, w can be defined as an integration of training patterns of x linearly. Therefore, Equation (2)
can be reformulated as [39]: N

 (x X
 x )( y
N
l l y)
R f (x) =l 1b + (αl − α∗l )hxl , xi (5)
N l=1 2 N (8)
where, αl , α∗l are the Lagrangian multipliers.
l 1
l (x
Then, in
x )l
order
l 1
(y  y )2
to put the nonlinearity in the algorithm,
the training patterns xl can be modified by a map Φ : χ → F . In addition, Kernel function can be
defined as: (9)
1 N
RMSE
N

k(x, x0 ) := Φ(x()x, lΦ(xy0 )l )
l 1
2
(6)

According to the above-mentioned equations, the optimization problem of (4) can be modified,
and finally, the function f derived as follows: 1 N
MSE 
NX
Nl 1
 (x l  y l )2
(10)
f (x) = b + (αl − α∗l )k(xl , x) (7)
l=1

N
1 problem, the flatness function is searched(11)

It is notable that in the nonlinear optimization among
MAE  xl  yl
the feature space, not input space [40,41]. N l 1
where x l and y l illustrate the actual value and predicted value. x and y y depict the mean of
actual values and forecasted values, respectively. Figure 3 shows the good correlation coefficient be‐
tween the real values and the predicted value by the network in the training, testing and validation
steps for the MLP network. Figure 4 indicates an excellent correlation coefficient between the real
values and the predicted value for the SVR network during the training phase.
Appl. Sci. 2020, 10, 3829 6 of 12

4. Simulation and Results

The MLP and SVR networks are designed to predict the cooling and heating load. Each of these
networks was trained using a dataset as input. In this paper, 85% (658 samples) of the data were used
to train and validate the proposed methods and the remaining 15% (110 samples) were used for testing.
In the first stage, each network required a preliminary design to determine the number of neurons
in the hidden layer and the coefficients of the network. After designing each network, the amount
of training and testing data for the network was determined. In this work, 70% of the samples were
considered as training data and 30% as test data to validate the training phase of each network. After
training and testing each neural network or regression algorithm, the results need to be evaluated.
To do this, correlation coefficient (R), mean squared error (MSE), root mean squared error (RMSE),
and mean absolute error (MAE) can be used as statistical performance metrics. Each of the mentioned
indices are calculated according to the following equations [42].

N
P
(xl − x)( yl − y)
l=1
R= s (8)
N N
2P
( yl − y)2
P
(xl − x)
l=1 l=1

v
u
N
t
1X
RMSE = (xl − yl )2 (9)
N
l=1

N
1X
MSE = (xl − yl )2 (10)
N
l=1
N
1X
MAE = xl − yl (11)
N
l=1

where xl and yl illustrate the actual value and predicted value. x and y y depict the mean of actual
values and forecasted values, respectively. Figure 3 shows the good correlation coefficient between the
real values and the predicted value by the network in the training, testing and validation steps for the
MLP network. Figure 4 indicates an excellent correlation coefficient between the real values and the
predicted value for the SVR network during the training phase.
Given the excellent correlation between the target data and the output of each network (as shown
in Figures 3 and 4), it can be clearly understood that each of these networks have passed the training
phase well. Good training means that the network is able to identify inherent patterns in the nature of
data and to predict the unknown data by using the learned patterns, so that each network learns how
much of a cooling and heating load is required for each building with specific characteristics. With this
training, each network can predict the amount of cooling and heating loads related to the input data of
the test phase. After training, each network is validated using initial test data (30% of 85%). This is
kind of a test for the training phase, which is done by the network itself. The prediction error in the
test or validation, which is one of the most important values in evaluating the results, is presented in
Figures 5 and 6 for each of the MLP and SVR networks in the histogram form, respectively.
The error obtained in the error histogram model indicates the minimum and maximum prediction
error. This means that in predicting the cooling and heating loads for the test data, the amount of error
that each of the trained networks can have is equal to the amount provided in the above figures.
In evaluating and analyzing each of the above figures that somehow represent the performance of
each network in the initial training and testing stages, it can be concluded that the training of proposed
methods has been well validated using the desired data. It should be noted that when the network is
trained with high accuracy, it is well designed and the amount of error in the validation and initial
Appl. Sci. 2020, 10, 3829 7 of 12

testing process depends more on the type of data. It also implies that the network will be able to
accurately assess and predict new and unknown data. Each network is saved as a black box after
training. This black box contains patterns that the network was able to identify during the training
phase. Now, the new and unknown data must be used to test these networks and predict the cooling
and heating load of buildings. To do this, 15% (110 samples) of the data, kept as unknown and new
data, were used. Figures 7 and 8 show the results of forecasting heating and cooling loads for new data
using
Appl. Sci.the trained
2020, 10, 3829MLP and SVR networks, respectively. 7 of 13
Appl. Sci. 2020, 10, 3829 7 of 13

(a)
(a) (b)
(b)
Figure
Figure 3.
3. Correlation
Correlationcoefficient
Correlation coefficientbetween
coefficient betweenthe
between thereal
the realvalue
real valueand
value andthe
and theoutput
the outputof
output ofthe
of theMLP:
the MLP:(a)
MLP: (a)heating
(a) heatingload;
heating load;
load;
(b)
(b) cooling
cooling load.
load.

(a)
(a) (b)
(b)
Figure
Figure 4.
4. Correlation
Correlationcoefficient
Correlation coefficientbetween
coefficient betweenthe
between thereal
the realvalue
real valueand
value andthe
and theoutput
the outputof
output ofthe
of theSVR;
the SVR;(a)
SVR; (a)heating
(a) heatingload;
heating load;
load;
(b) cooling load.
(b) cooling load.

Given
Given the
the excellent
excellent correlation
correlation between
between thethe target
target data
data andand the
the output
output ofof each
each network
network (as (as
shown in Figures 3 and 4), it can be clearly understood that each of these networks have
shown in Figures 3 and 4), it can be clearly understood that each of these networks have passed the passed the
training
training phase
phase well.
well. Good
Good training
training means
means that
that the
the network
network isis able
able to
to identify
identify inherent
inherent patterns
patterns in
in the
the
nature
nature of
of data
data and
and to
to predict
predict the
the unknown
unknown data
data by
by using
using the
the learned
learned patterns,
patterns, so
so that
that each
each network
network
learns
learns how
how much
much of
of aa cooling
cooling and
and heating
heating load
load is
is required
required for
for each
each building
building with
with specific
specific character‐
character‐
Appl. Sci. 2020, 10, 3829 8 of 13

results, is presented
presented in Figures
5 and56and
for 6each
for each of MLP
the MLP
and and
SVR SVR networks in histogram
the histogram
form,form,
Appl. Sci. 2020, 10, in Figures
3829 of the networks in the 8 of 12
respectively.
respectively.

(a) (b)
(a) (b)
Figure 5. MLP testing error in the for histogram; (a) heating load; (b) cooling load.
Figure 5. MLP
MLP testing
testing error
error in the for histogram; (a) heating load; (b) cooling load.

(a) . (b)
Appl. Sci. 2020, 10, 3829 (a) (b) 9 of 13
Appl. Sci. 2020, 10,
Figure 6. SVR testing error in the for histogram; (a) heating load; (b) cooling load. 9 of 13
3829 6. SVR testing error in the for histogram; (a) heating load; (b) cooling load.
Figure
Figure 6. SVR testing error in the for histogram; (a) heating load; (b) cooling load.
The error obtained in the error histogram model indicates the minimum and maximum prediction
error. This means that in predicting the cooling and heating loads for the test data, the amount of
The error obtained in the error histogram model indicates the minimum and maximum predic‐
error that each of the trained networks can have is equal to the amount provided in the above figures.
tion error. This means that in predicting the cooling and heating loads for the test data, the amount
In evaluating and analyzing each of the above figures that somehow represent the performance of
of error that each of the trained networks can have is equal to the amount provided in the above
each network in the initial training and testing stages, it can be concluded that the training of
figures.
proposed methods has been well validated using the desired data. It should be noted that when the
In evaluating
network andwith
is trained analyzing each ofitthe
high accuracy, above
is well figuresand
designed that thesomehow
amount ofrepresent thevalidation
error in the performance
of each
andnetwork in the
initial testing initial
process training
depends moreand ontesting
the typestages,
of data.itIt can
also be concluded
implies that the that the will
network training
be of
proposed
able tomethods
accurately has beenand
assess well validated
predict new andusing the desired
unknown data.network
data. Each It should be noted
is saved as a that
blackwhen
box the
afteristraining.
network trainedThiswithblack
highbox containsit patterns
accuracy, that the network
is well designed and thewas able to
amount of identify
error induring the
the validation
and initial testing process depends more on the type of data. It also implies that the networkthe
training phase. Now, the new and unknown data must be used to test these networks and predict will be
cooling
able to and heating
accurately assessload
andofpredict
buildings.
new Toanddo this, 15% (110
unknown samples)
data. of the data,
Each network is kept
savedas as
unknown
a black box
and new data, were used. Figures 7 and 8 show the results of forecasting heating and cooling loads
after training. This black box contains patterns that the network was able to identify during the train‐
for new data using the trained
Figure 7.MLP and load
Heating SVRforecasting
networks,via respectively.
MLP and SVR.
ing phase. Now, the new Figure Figure Heating
7.7.Heating
and unknown load
data
load forecasting
must
forecastingbevia via
used
MLP MLP
to
andtestandthese
SVR. SVR. networks and predict the
cooling and heating load of buildings. To do this, 15% (110 samples) of the data, kept as unknown
and new data, were used. Figures 7 and 8 show the results of forecasting heating and cooling loads
for new data using the trained MLP and SVR networks, respectively.

Figure Coolingload
8.Cooling
Figure 8. load forecasting
forecasting via via
MLPMLP and SVR.
and SVR.
Figure 8. Cooling load forecasting via MLP and SVR.
Performance evaluation of the proposed methods is presented in Table 2 in terms of R, MSE,
Performance evaluation of the proposed methods is presented in Table 2 in terms of R, MSE,
RMSE, and MAE.
RMSE, and MAE.
Table 2. Results of accuracy and error for proposed methods in heating and cooling load prediction.
Table 2. Results of accuracy and error for proposed methods in heating and cooling load prediction.
Heating load Cooling load
Heating load Cooling load
R MSE RMSE MAE R MSE RMSE MAE
R MSE RMSE MAE R MSE RMSE MAE
Appl. Sci. 2020, 10, 3829 9 of 12

Performance evaluation of the proposed methods is presented in Table 2 in terms of R, MSE,

RMSE, and MAE.

Table 2. Results of accuracy and error for proposed methods in heating and cooling load prediction.

Heating Load Cooling Load

R MSE RMSE MAE R MSE RMSE MAE
MLP 0.9993 0.2335 0.4832 0.4118 0.9824 6.896 2.626 2.0973
SVR 0.9979 0.7838 0.8853 0.7780 0.9878 3.024 1.7389 1.4762

Based on the results presented in the Table 2, it can be seen that the best prediction was related
to the prediction of the heating load by the MLP method with the highest value of R (0.9993) and
minimal errors in the form of MSE (0.2335), RMSE (0.4832), and MAE (0.4118). However, in predicting
the cooling load, the SVR method with a large amount of R (0.9878) and lowest errors in the terms of
MSE (3.024), RMSE (1.7389), and MAE (1.4762) made a good prediction. Highest values of MSE and
RMSE errors of prediction were also related to the MLP method in the prediction of cooling load. The
use of machine learning methods and their results are highly dependent on the type of input data.
It is observed that there is a difference between the results of predicting the cooling load and heating
load by each of the networks and the heating load is predicted with high accuracy. This difference
is due to the poor correlation between the input data and the amount of cooling load relative to the
heating load. To evaluate the effectiveness of the proposed methods in this paper, it is necessary to
compare the results obtained with the results of other studies. Comparisons should be made with
caution using similar datasets. To this end, a number of studies were selected for comparison in which
similar data was used for predicting the cooling and heating loads. To express the effectiveness of
the data type in the accuracy of the results, the results of several studies conducted to predict cooling
and heating loads using different data were compared with the results obtained in this paper. Table 3
makes this comparison.

Table 3. Comparison of cooling and heating loads prediction results with other works.

Data Type References Heating Load (R) Cooling Load (R)

Used data in this paper MLP in this paper 0.9993 0.9824
SVR in this paper 0.9979 0.9878
DNN [14] 0.9805 0.9976
GBM [14] 0.9853 0.9853
GPR [14] 0.9984 0.9913
MPMR [14] 0.8802 0.8955
ANN [15] 0.9980 0.9840
CART [15] 0.9960 0.9810
GLR [15] 0.9950 0.9830
CHAID [15] 0.9950 0.9810
GA-ANN [18] 0.9800 -
PSO-ANN [18] 0.9720 -
ICA-ANN [18] 0.9700 -
ABC-ANN [18] 0.9730 -
Different data GRNN [28] - 0.9640
PENN [20] - 0.9500
MLR [20] - 0.7510
AR [20] - 0.8370
ARX [20] - 08640
MNR (initial prediction) [20] - 0.8990
MNR (final calibration) [20] - 0.9580
ANN [21] 0.9900 -
Decision tree [22] 0.92 -
Appl. Sci. 2020, 10, 3829 10 of 12

The comparison made in Table 3, shows the accuracy and efficiency of the proposed methods
in this paper for forecasting the cooling and heating loads of the building. The use of machine
learning applications and the selection of the applicable method for energy predicting and energy
efficiency measures in residential buildings are significantly effective in saving energy consumption.
The selected methods were able to realize the purpose of the paper with their high accuracy and
achieve this important goal. Finally, it should be noted that the proposed solutions can also be used for
real-world data.

5. Conclusions
Nowadays, the importance of energy saving and its management has raised many challenges
in forecasting the heating and cooling loads of buildings. Most researchers in this field offer many
methods and models for predicting heating and cooling loads to somehow increase the prediction
accuracy. In this paper, based on machine learning models two MLP and SVR methods were proposed
to predict the cooling and heating load of a residential building. The main idea of these methods was
to create a linear mapping between the input and output variables to increase the prediction accuracy.
After designing each of the proposed models, the technical parameters of a home building were used as
inputs and the heating and cooling loads were used as the output variables of each network during the
training phase. New and anonymous data were used to test the trained networks and for forecasting
the heating and cooling loads. Finally, each trained network was able to reliably provide the heating
and cooling load forecasts. Meanwhile, the MLP method with the maximal of R i.e., 0.9993 and the SVR
method with the highest value of R i.e., 0.9878 predicted the heating and cooling loads, respectively.

Author Contributions: A.M.: Writing—original draft, software, methodology, and validation: A.M.-S.; formal
analysis, investigation, resources, writing—review and editing: B.M.-I.; conceptualization, data curation,
writing—review and editing: A.A.-M.; supervision, writing—review and editing. All authors have read
and agreed to the published version of the manuscript.
Funding: This research was supported by the HeatReFlex-Green and Flexible Heating/Cooling project, (www.
heatreflex.et.aau.dk) funded by Danida Fellowship Centre and the Ministry of Foreign Affairs of Denmark under
the grant no. 18-M06-AAU.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Dadon, I. Planning the Second Generation of Smart Cities: Technology to handle the pressures of urbanization.
IEEE Electrif. Mag. 2019, 7, 6–15. [CrossRef]
2. Pérez-Lombard, L.; Ortiz, J.; Pout, C. A review on buildings energy consumption information. Energy Build.
2008, 40, 394–398. [CrossRef]
3. Moradzadeh, A.; Sadeghian, O.; Pourhossein, K.; Mohammadi-Ivatloo, B.; Anvari-Moghaddam, A. Improving
Residential Load Disaggregation for Sustainable Development of Energy via Principal Component Analysis.
Sustainability 2020, 12, 3158. [CrossRef]
4. Wei, Y.; Zhang, X.; Shi, Y.; Xia, L.; Pan, S.; Wu, J.; Han, M.; Zhao, X. A review of data-driven approaches
for prediction and classification of building energy consumption. Renew. Sustain. Energy Rev. 2018, 82,
1027–1047. [CrossRef]
5. Abd Alla, S.; Bianco, V.; Tagliafico, L.A.; Scarpa, F. Life-cycle approach to the estimation of energy efficiency
measures in the buildings sector. Appl. Energy 2020, 264, 114745. [CrossRef]
6. Zhao, J.; Liu, X. A hybrid method of dynamic cooling and heating load forecasting for office buildings based
on artificial intelligence and regression analysis. Energy Build. 2018, 174, 293–308. [CrossRef]
7. Di Foggia, G. Energy efficiency measures in buildings for achieving sustainable development goals. Heliyon
2018, 4, e00953. [CrossRef]
8. Shi, E.; Jabari, F.; Anvari-Moghaddam, A.; Mohammadpourfard, M.; Mohammadi-Ivatloo, B.
Risk-Constrained Optimal Chiller Loading Strategy Using Information Gap Decision Theory. Appl. Sci. 2019,
9, 1925. [CrossRef]
Appl. Sci. 2020, 10, 3829 11 of 12

9. Nebot, À.; Mugica, F. Energy performance forecasting of residential buildings using fuzzy approaches.
Appl. Sci. 2020, 10, 720. [CrossRef]
10. Yang, L.; Yan, H.; Lam, J.C. Thermal comfort and building energy consumption implications—A review.
Appl. Energy 2014, 115, 164–173. [CrossRef]
11. Mansour-Saatloo, A.; Agabalaye-Rahvar, M.; Mirzaei, M.A.; Mohammadi-Ivatloo, B.; Zare, K. Robust
scheduling of hydrogen based smart micro energy hub with integrated demand response. J. Clean. Prod.
2020, 267, 122041. [CrossRef]
12. Mirzaei, M.A.; Yazdankhah, A.S.; Mohammadi-Ivatloo, B. Stochastic security-constrained operation of wind
and hydrogen energy storage systems integrated with price-based demand response. Int. J. Hydrogen Energy
2019, 44, 14217–14227. [CrossRef]
13. Jozi, A.; Pinto, T.; Praça, I.; Vale, Z. Decision support application for energy consumption forecasting.
Appl. Sci. 2019, 9, 699. [CrossRef]
14. Zhou, Y.; Wu, J.; Long, C. Evaluation of peer-to-peer energy sharing mechanisms based on a multiagent
simulation framework. Appl. Energy 2018, 222, 993–1022. [CrossRef]
15. Cui, S.; Wang, Y.-W.; Xiao, J.-W. Peer-to-Peer Energy Sharing among Smart Energy Buildings by Distributed
Transaction. IEEE Trans. Smart Grid 2019, 10, 6491–6501. [CrossRef]
16. Kneifel, J. Life-cycle carbon and cost analysis of energy efficiency measures in new commercial buildings.
Energy Build. 2010, 42, 333–340. [CrossRef]
17. Penna, P.; Prada, A.; Cappelletti, F.; Gasparella, A. Multi-objectives optimization of Energy Efficiency
Measures in existing buildings. Energy Build. 2015, 95, 57–69. [CrossRef]
18. Macas, M.; Moretti, F.; Fonti, A.; Giantomassi, A.; Comodi, G.; Annunziato, M.; Pizzuti, S.; Capra, A. The
role of data sample size and dimensionality in neural network based forecasting of building heating related
variables. Energy Build. 2016, 111, 299–310. [CrossRef]
19. Jing, Z.; Cai, M.; Pipattanasomporn, M.; Rahman, S.; Kothandaraman, R.; Malekpour, A.; Paaso, E.A.;
Bahramirad, S. Commercial Building Load Forecasts with Artificial Neural Network. In Proceedings of the
2019 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), Washington, DC,
USA, 18–21 February 2019; pp. 1–5.
20. Zhang, L.; Wen, J. A systematic feature selection procedure for short-term data-driven building energy
forecasting model development. Energy Build. 2019, 183, 428–442. [CrossRef]
21. Le, L.T.; Nguyen, H.; Dou, J.; Zhou, J. A comparative study of PSO-ANN, GA-ANN, ICA-ANN, and
ABC-ANN in estimating the heating load of buildings’ energy efficiency for smart city planning. Appl. Sci.
2019, 9, 2630. [CrossRef]
22. Kwok, S.S.K.; Lee, E.W.M. A study of the importance of occupancy to building cooling load in prediction by
intelligent approach. Energy Convers. Manag. 2011, 52, 2555–2564. [CrossRef]
23. Fan, C.; Ding, Y. Cooling load prediction and optimal operation of HVAC systems using a multiple nonlinear
regression model. Energy Build. 2019, 197, 7–17. [CrossRef]
24. Chaudhuri, T.; Soh, Y.C.; Li, H.; Xie, L. A feedforward neural network based indoor-climate control framework
for thermal comfort and energy saving in buildings. Appl. Energy 2019, 248, 44–53. [CrossRef]
25. Yu, Z.; Haghighat, F.; Fung, B.C.M.; Yoshino, H. A decision tree method for building energy demand
modeling. Energy Build. 2010, 42, 1637–1646. [CrossRef]
26. Roy, S.S.; Samui, P.; Nagtode, I.; Jain, H.; Shivaramakrishnan, V.; Mohammadi-ivatloo, B. Forecasting heating
and cooling loads of buildings: A comparative performance analysis. J. Ambient Intell. Humaniz. Comput.
2020, 11, 1253–1264. [CrossRef]
27. Chou, J.-S.; Bui, D.-K. Modeling heating and cooling loads by artificial intelligence for energy-efficient
building design. Energy Build. 2014, 82, 437–446. [CrossRef]
28. Bacher, P.; Madsen, H.; Nielsen, H.A.; Perers, B. Short-term heat load forecasting for single family houses.
Energy Build. 2013, 65, 101–112. [CrossRef]
29. Yu, J.; Bae, M.; Bang, H.-C.; Kim, S.-J. Cloud-based building management systems using short-term cooling
load forecasting. In Proceedings of the 2013 IEEE Globecom Workshops (GC Wkshps), Atlanta, GA, USA,
9–13 December 2013; pp. 896–900.
30. Favre, L.; Schafer, T.M.; Robyr, J.-L.; Niederhäuser, E.-L. Intelligent algorithm for energy, both thermal and
electrical, economic and ecological optimization for a smart building. In Proceedings of the 2018 IEEE
International Energy Conference (ENERGYCON), Limassol, Cyprus, 3–7 June 2018; pp. 1–5.
Appl. Sci. 2020, 10, 3829 12 of 12

31. Ben-Nakhi, A.E.; Mahmoud, M.A. Cooling load prediction for buildings using general regression neural
networks. Energy Convers. Manag. 2004, 45, 2127–2141. [CrossRef]
32. Tsanas, A.; Xifara, A. Accurate quantitative estimation of energy performance of residential buildings using
statistical machine learning tools. Energy Build. 2012, 49, 560–567. [CrossRef]
33. Haykin, S.S. Neural Networks and Learning Machines/Simon Haykin; Prentice Hall: New York, NY, USA, 2009.
34. Moradzadeh, A.; Khaffafi, K. Comparison and evaluation of the performance of various types of neural
networks for planning issues related to optimal management of charging and discharging electric cars in
intelligent power grids. Emerg. Sci. J. 2017, 1, 201–207.
35. Moradzadeh, A.; Pourhossein, K. Early Detection of Turn-to-Turn faults in Power Transformer Winding:
An Experimental Study. In Proceedings of the 2019 International Aegean Conference on Electrical Machines
and Power Electronics (ACEMP) & 2019 International Conference on Optimization of Electrical and Electronic
Equipment (OPTIM), Istanbul, Turkey, 27–29 August 2019; pp. 199–204.
36. Seo, D.K.; Eo, Y.D. Multilayer Perceptron-Based Phenological and Radiometric Normalization for
High-Resolution Satellite Imagery. Appl. Sci. 2019, 9, 4543. [CrossRef]
37. Thimm, G.; Fiesler, E. High-order and multilayer perceptron initialization. IEEE Trans. Neural Netw. 1997, 8,
349–359. [CrossRef] [PubMed]
38. Moradzadeh, A.; Pourhossein, K. Application of Support Vector Machines to Locate Minor Short Circuits
in Transformer Windings. In Proceedings of the 2019 54th International Universities Power Engineering
Conference (UPEC), Bucharest, Romania, 3–6 September 2019; pp. 1–6.
39. Zhang, J.; Liao, Y.; Wang, S.; Han, J. Study on driving decision-making mechanism of autonomous vehicle
based on an optimized support vector machine regression. Appl. Sci. 2018, 8, 13. [CrossRef]
40. Smola, A.J.; Schölkopf, B. A tutorial on support vector regression. Stat. Comput. 2004, 14, 199–222. [CrossRef]
41. Taghavifar, H.; Mardani, A. A comparative trend in forecasting ability of artificial neural networks and
regressive support vector machine methodologies for energy dissipation modeling of off-road vehicles.
Energy 2014, 66, 569–576. [CrossRef]
42. Choubin, B.; Khalighi-Sigaroodi, S.; Malekian, A.; Kişi, Ö. Multiple linear regression, multi-layer perceptron
network and adaptive neuro-fuzzy inference system for forecasting precipitation based on large-scale climate
signals. Hydrol. Sci. J. 2016, 61, 1001–1009. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).