sustainability

Article
Occupant Comfort Management Based on Energy
Optimization Using an Environment Prediction
Model in Smart Homes
Wenquan Jin , Israr Ullah , Shabir Ahmad and Dohyeun Kim *
Department of Computer Engineering, Jeju National University, Jeju 63243, Korea;
[email protected] (W.J.); [email protected] (I.U.); [email protected] (S.A.)
* Correspondence: [email protected]




Received: 7 January 2019; Accepted: 10 February 2019; Published: 15 February 2019


Abstract: Occupant comfort management is an important feature of a smart home, which requires
achieving a high occupant comfort level as well as minimum energy consumption. Based on a large
amount of data, learning models enable us to predict factors of a mathematical model for deriving the
optimal result without expensive experiments. Comfort management supports high-level comfort
to the occupant in the individual indoor environment, using the optimal power consumption to
run home appliances. In this paper, we propose occupant comfort management based on energy
optimization, using an environment prediction model. The proposed energy optimization model
provides optimal power consumption based on the proposed objective function, which requires
temperature and comfort index data as the input parameters. For the input requirement, temperature
prediction model and humidity prediction model are presented based on a recurrent neural network
with a pre-collected dataset, including indoor and outdoor temperature and humidity sensing data.
Using the predicted temperature and humidity data, the comfort index model derives the predicted
mean vote value to be used in the energy optimization model with the predicted temperature data.
The experimental results present an 8.43% reduction of the optimized power consumption compared
to the actual power consumption using mean absolute percentage error to calculate. Moreover, the
emulation of an indoor environment using optimal energy consumption presents as approximately
similar to the actual data.

Keywords: prediction; recurrent neural networks (RNN); user comfort; predicted mean vote (PMV);
energy optimization; objective function

1. Introduction
In modern life, various electronic appliances provide convenience and improve people’s living
quality in a way that is based on the consumption of electronic energy. These electronic appliances
are one of the biggest consumers of electricity in many cities, where people live urban lives with
many electronic products day and night [1]. Especially in buildings, a huge amount of electricity in
the form of final electronic energy is consumed, which is a larger proportion than both industry and
transportation in many developed countries [2]. As shown through the survey, people spend more
than 80% of their days in the indoor environment, using electronic products for living, working, and
entertainment [3]. With the development of internet infrastructures, much work will be done through
the remote access by employees in the indoor environment, such as their home, office, and wherever
people feel comfor working. Furthermore, according to the United Nations, the population of urban
centers will increase from today’s 55% to 68% in 2050, including most people in developing countries [4].
This growth is leading to increasing energy consumption by the operation of electronic appliances

Sustainability 2019, 11, 997; doi:10.3390/su11040997 www.mdpi.com/journal/sustainability

Sustainability 2019, 11, 997 2 of 18

in the buildings. Nevertheless, providing a comfortable environment to occupants is an important

function of the indoor environment because of the occupant’s living quality, including the health and
productivity [5–8]. Therefore, in any building, an intelligent approach is required for maintaining the
user’s environmental comfort while consuming minimal energy through electronic appliances.
The intelligent control system of buildings is comprised of a computer, data storage, sensor, and
actuator, to support the occupant’s expected comfort while reducing energy consumption during the
operation of building [9]. For supporting a comfortable environment, the intelligent control system
requires parameters, including various indoor and outdoor environmental sensing data with user
preferences, to build the model. The parameters are difficult to collect for the mathematical solution,
which may require the parameters in specific cases for previous studies [10–12]. The limitations of
deploying various sensors to satisfy the condition of the formula for calculating the comfort index
with minimal energy consumption is a challenge [13]. The comfort index can be predicted through a
historical dataset for the individual environment [14]. Predicting various parameters is important to
get the factors for making a comfortable environment for users without human intervention, where
the electronic appliances work by consuming the power. Through prediction, outputs can be used
to change the environment through the actuators. The environmental parameters are the inputs of
the prediction model to affect the factors of a control system for heating, cooling, ventilation, and
lighting [15]. In order to minimize power consumption with the comfortable environment, the control
system of smart buildings is required to include the functionalities for supporting high-level comfort,
high power efficiency, and user-friendliness.
However, the tradeoff between the energy consumption and indoor comfort level is difficult to
achieve, because energy consumption depends on the total operational cost of electronic appliances,
and the comfort experience is affected by various dynamic factors, including the occupant’s state in
the conditioned environment [16]. Furthermore, for supporting comfort, various electronic appliances
collaborate, such as heating, ventilation, and air conditioning (HVAC), gas, and lighting [17,18].
Therefore, supporting the comfortable environment continuously without human intervention shall
consume unnecessary energy for satisfying the conditions of user preference. A simple rule-based
control system can adjust these actuators based on the collected sensing data from the sensors for each
environmental parameter [19,20]. Therefore, frequent controlling is required for changing the desired
temperature, humidity, and other environmental parameters through the actuators. However, the
frequent operation of actuators results in consumption of more energy.
In this paper, we propose a scheme of imposed energy optimization for supporting the high-level
comfort to the user in the individual indoor environment using optimal power consumption.
The indoor environment is comprised of sensors, actuators, indoor environmental parameters, and the
user. For the awareness of the environment, various environmental data is required to be collected.
The data on outdoor environmental parameters can be collected from the weather station, and the
data of indoor environmental parameters can be collected from indoor sensors. Using the collected
data of environmental parameters, the environment can be predicted based on prediction models.
The recurrent neural network (RNN) is a machine learning (ML) algorithm that is used for forecasting
parameters in the time sequence based on involving the initial states and the past states of the neurons
in the series of processing [21]. For predicting the indoor environment, an RNN-based prediction
model is used occupant comfort management. The prediction model forecasts the indoor environment
for the user using the individual historical dataset, including indoor and outdoor data, which is
collected by the Oak Ridge National Laboratory (ORNL) to be published online [22]. The predicted
indoor environment in the future is used for calculating the comfort index and supporting parameters
of the objective function in the optimizer. The comfort index is derived by the predicted mean vote
(PMV) formula, which provides physical phenomenon and an outcome of human thermal sensation for
the heat transfer between a human body and its surrounding environment [23]. The objective function
is implemented using a complex formula from an optimal control strategy used for a multi-zone air
conditioning system [24]. The control factor is inferred by the objective function of the optimizer, based
Sustainability 2019, 11, 997 3 of 18

on the future comfort index and other environmental parameters. Through comfort management,
including a prediction model and optimizer, the optimal power consumption is applied to change the
indoor environment for achieving maximum user comfort with minimum energy consumption.
The rest of the paper is structured as follows. Section 2 introduces related works regarding
occupants’ building environment control systems, based on advanced approaches of fuzzy logic,
optimization, and ML. Section 3 illustrates the proposed occupant comfort management, including
models of environment parameter prediction, the comfort index, and energy consumption optimization.
Section 4 presents prediction and optimization results regarding indoor environmental parameters
and energy consumption. Section 5 presents our conclusions regarding the proposed occupant comfort
management model.

2. Related Works
Energy consumption is of paramount importance in residual buildings, due to the fact that most
of the buildings consume energy to ensure a certain degree of comfortable environment for occupants.
Controlling temperature in an individual space for residents is one of the fundamental functions in
occupant buildings. In literature, numerous comfort indices have been considered, including primarily
thermal comfort, visual comfort, and air quality. These indices are the function of environmental
parameters, such as temperature, humidity, illumination, and air quality [25]. Numerous comfort
management strategies have been devised to exploit the indoor and outdoor environmental parameters
and ensure a certain degree of comfort for occupants in these buildings. In [26–28], an optimal model
is proposed to provide a solution for energy saving, and to predict the factors of heating, ventilating,
and air conditioning systems. Many recent studies leverage fuzzy logic controllers to control the
environmental parameters through actuators [29,30]. These methods primarily deal with single
object optimization, and thus have addressed a single-factor controlling mechanism in the comfort
management system. However, in reality such systems require multiple factors to ensure a comfortable
environment. Model predictive control (MPC) scheme is another approach, in which the optimal
control can be reached through modeling the future environment based on the operation strategy for
saving energy and making a comfortable environment [31]. A variation of this approach has been
proposed, which introduces feedback MPC and allows automatic control in a building in order to
maintain energy efficiency and the self-operation of building actuators [32]. Weighted parameters in
the objective function have been employed to get the optimal control factor, in order to solve multiple
parameters for the complex environment [33–35].
For achieving optimal values of various parameters in the objective function, an optimization
algorithm is required to be selected. Many optimization algorithms have been proposed that are
inspired from the behavior of nature. These optimization algorithms are modelled to get optimal
results based on reaching the optimal value in each parameter of the objective function. For instance,
colony optimization uses stochastic nature to reach its optimal value [36]. Similarly, the genetic
algorithm (GA) and particle swarm optimization (PSO) are evolutionary algorithms, which can be
used for non-linear optimization problems to achieve optimization [37]. GA and PSO are used in
many different cases to optimize power consumption while controlling electronic appliances [38–40].
Certain instances from the literature have also been found where GA is utilized in the building control
system, in order to optimize the control factor of HVAC to reach the optimal performance [41]. PSO,
on the other hand, is used in the recommendation system of buildings to minimize the annual energy
consumption for the building energy performance optimization [42]. These optimization solutions
provide a real-time solution for complex multi-objective optimization problems, which most linear
mathematical models cannot realistically solve. Based on these algorithms, objective functions are
proposed to give the optimal values in the comfort management system. However, these objective
functions require multiple parameters in the complex environment [43,44]. In the proposed comfort
management scheme, the parameters are inferred through the prediction model and calculated by
the PMV formula. Heater power consumption is the only value to be required for achieving optimal
Sustainability 2019, 11, 997 4 of 18

comfort from the objective function. Therefore, optimization algorithms are not required to optimize
the objective function in the proposed scheme.
Energy consumption has become increasingly important to control the stress on today’s electrical
grid infrastructure [45]. Therefore, predicting energy consumption is crucial to find energy demand in
the future, for the reliability of power supply system. Through the prediction model, the parameters
can be based on the inputs of the model. For forecasting of time series, historical data are required
as the inputs of the model, to predict future data in the time sequence. A variety of ML algorithms
have been proposed for forecasting, such as the artificial neural network (ANN), hidden Markov
model (HMM), support vector machine (SVM), and many other ML algorithms [46]. The ANNs
have been used for solving nonlinear methods, which can be trained to learn the relationship for
recognizing patterns from given data, through the functions of self-organizing, data-driven, self-study,
self-adaptive, and associated memory [47]. For the training algorithms to predict the parameter
in a time sequence, various neural network architectures have been proposed, such as generalized
regression, wavelet, dynamic, and backpropagation [48]. Recently, most of the neural networks have
been based on multiple layers for enabling the complex mapping between inputs and outputs, in order
to learn the approximate nonlinear relationship [49]. Multi-layer neural networks are based on two
basic architectures: convolutional neural network (CNN) and RNN [50]. The CNN is handy in cases
like image recognition, where the image data is independent from the time series [51,52].
In contrast, RNN contains a memory component, to pass the selected information to the outputs
of the model for capturing the time-related dependencies [53]. Using an RNN based prediction model,
the long short-term memory (LSTM) can be used for learning time series with long time spans [54].
Therefore, for forecasting the data continued in the time series, the model of an RNN is trained by
the continuous data to pass the features on to the next time in the timeline. For forecasting wind
speed, Cheng et al. present an RNN-based architecture to maintain diversities of sub-models and
take advantages of learning [55]. For estimating the charging state of an Li-ion battery without using
battery models, filters, and inference systems like Kalman filters, Chemali et al. introduce an RNN
with an LSTM model to predict the charging state, based on continuous battery life [56]. Microsoft
engineers Tam et al. proposed an RNN-based, spoken-language understanding model to generate
valid unseen labeled data, through capturing the structure of a word sequence for evaluating live logs
of Cortana [57]. For a short-term traffic data forecast in the transportation system, Zhao et al. introduce
an RNN-based data prediction approach using LSTM cells [58]. Therefore, most of the forecasting
problems are implemented by an RNN that can deliver the information from data to the next data that
are continuous in the time sequence.
For user comfort, the PMV is the most widely used model, and was developed through costly
laboratory experiments by Fanger in 1970s [59]. The PMV model is the basis of the International
Organization for Standardization (ISO) 7730 standard, which has 1994 and 2005 versions [60,61].
A personal comfort model results in individual thermal comfort, which presents the comfort level
based on the person’s surrounding environment. The PMV-based comfort model takes an individual
person as the unit instead of the average of a group, and uses direct feedback from individuals,
including thermal sensation, preference, acceptability and pleasure. A personal comfort model is
used for better understanding regarding specific comfort needs and the desired thermal comfort of an
occupant in an individual environment. For intelligent comfort management, the comfort data can
be used for the intelligent controlling model, to manage the building or other environments and to
achieve optimal conditioning with energy efficiency [62–64]. In the proposed comfort management,
the PMV results from the predicted environment for the future comfort of the individual occupant
environment, to be used for optimizing power consumption within the comfortable environment.

3. Proposed Occupant Comfort Management Methodology

For occupant comfort management in smart spaces, where the sensors and actuators are deployed
to support a comfortable environment to the user, the intelligent system needs to collect environmental
The comfort index is used for evaluating the indoor environment, where the user stays. Through the
comfort index, the system can make a strategy to control the actuators for updating the indoor
environmental parameters to support the comfortable environment. The comfort index calculator is
a generator for the comfort index, which outputs the data for evaluating the indoor environment.
The optimizer
Sustainability 2019, 11, 997includes the objective function for optimizing the power parameter for the
5 of 18
environment using environment parameters, including historical data, predicted data, and comfort
index. The output of the optimizer is a power consumption value that is used for operating the
information
actuator and make
to change the smart
the indoor decision for
environment for controlling the actuators
the user comfort. to results
The value changefrom
the environment,
the objective
using appropriate power consumption. Figure 1 presents the power consumption
function for the minimum power consumption of the comfortable environment. Therefore, optimization process,
the
which is based
actuators on environment
shall make prediction.
the environment comfortable for the user using optimal power consumption.

Environment Prediction Model Comfort Index Model

Predicted Indoor
Environmental Data
Comfort Index

Energy Optimization Model

Objective Function
Outdoor Environmental
Data

Indoor Environmental Data Power Parameter

Weather Station Sensors Actuators

Outdoor Environmental Indoor Environmental

Parameters parameters

Outdoor Environment Indoor Environment

Figure 1.
Figure Energy consumption
1. Energy consumption optimization
optimization based
based on
on environmental
environmental parameter
parameter prediction.
prediction.

The prediction
3.1. Environmental model requires
Parameter Predictiona set
Modelof data that is used for predicting the next time. In order
to develop the prediction model, the dataset is used for training the model. The prediction model is
used For predicting the
for predicting theindoor
environmental parameters,
environment, which isprediction
comprisedmodels are environmental
of various proposed for temperature
parameters.
dataparameters
The and humidity data in
change thetime.
over indoor space. The
Therefore, thepredicted
data must data is used for calculating
be continuous for trainingthe
thePMV value
model, as
that is the comfort index, to present user comfort in the future indoor environment. The
well as predicting the next time. The input data that is used for predicting the next time, which can beproposed
optimizer
a data momentalso requires
or a data the predicted
sequence fromindoor
a pointenvironmental datatowith
in time in the past the future
the current comfort
point. index to
The outputs of
infer the optimal power. Figure 2 shows the prediction models to predict the temperature
the prediction model are a dataset for the predicted indoor environmental parameters in the future. data and
humidity
This data, using
forecasting historical
is continued data
from forpast,
the indoor andisoutdoor
which affectedenvironmental
by all the data parameters.
from the past.
Using the outputs of the prediction model, the comfort index can be calculated for the future.
The comfort index is used for evaluating the indoor environment, where the user stays. Through
the comfort index, the system can make a strategy to control the actuators for updating the indoor
environmental parameters to support the comfortable environment. The comfort index calculator is a
generator for the comfort index, which outputs the data for evaluating the indoor environment.
The optimizer includes the objective function for optimizing the power parameter for the
environment using environment parameters, including historical data, predicted data, and comfort
index. The output of the optimizer is a power consumption value that is used for operating the actuator
to change the indoor environment for the user comfort. The value results from the objective function
for the minimum power consumption of the comfortable environment. Therefore, the actuators shall
make the environment comfortable for the user using optimal power consumption.
Sustainability 2019, 11, x FOR PEER REVIEW 6 of 18
Sustainability 2019, 11, 997 6 of 18

The proposed prediction models require input parameters, such as time sequence (TS), outdoor
temperature (To), Parameter
3.1. Environmental user-desired indoorModel
Prediction temperature (Ti), outdoor humidity (Ho), and user-desired
indoor humidity (Hi). The temperature prediction model (TPM) requires historical data for the
For predicting
environmental the environmental
parameters To and Ti.parameters, prediction models
For each environmental are proposed
parameter, for temperature
the TPM requires 96
data and humidity data in the indoor space. The predicted data is used for calculating the PMV value
sequential data. Each row of the input dataset is comprised of a TS, To, and Ti. The 96 sequential data
that is the comfort index, to present user comfort in the future indoor environment. The proposed
are the historical data from t-96 to t. Each t represents the number of TSs that present 15 minutes. The
optimizer
t means the also requires
number for the
the predicted
current timeindoor
in theenvironmental
TS. Therefore, data
96 ofwith the future
TS presents comfort
1 hour. indexthe
Through to
infer the optimal power. Figure 2 shows the prediction models to predict the temperature
TPM, the Ti of t + 1 can be predicted. Similarly, the humidity prediction model (HPM) requires the data and
humidity
historical data, using
data of Ho andhistorical
Hi. Thedata for indoor
output of the and
HPM outdoor environmental
is the Hi of t + 1. parameters.

Historical Data Predicted Data

(TS, To, Ti)
Temperature
(TS, To, Ti)
Prediction (Ti)
(TS, To, Ti)
Model
…
(TS, To, Ti)
(TS, To, Ti)
(TS, To, Ti)

(TS, Ho, Hi) Humidity

(TS, Ho, Hi) Prediction (Hi)
(TS, Ho, Hi) Model
…
(TS, Ho, Hi)
(TS, Ho, Hi) TS: Time Sequence
(TS, Ho, Hi) To: Outdoor Temperature
Ti: User Desired Indoor Temperature
Ho: Outdoor Humidity
Hi: User Desired Indoor Humidity

Environmental prediction model based on indoor and outdoor temperature and humidity.
Figure 2. Environmental humidity.

The
The proposed
predictionprediction
models are models
built byrequire input parameters,
the RNN-based ML, which suchisas timeinsequence
good scenarios(TS), outdoor
where there
temperature
are temporal (To), user-desired
or order dependencies indoor temperature
through containing(Ti),aoutdoor
memoryhumidity
component, (Ho), and user-desired
to pass the selective
indoor
knowledge humidity
down the(Hi).time The temperature
sequence. Figureprediction
3 shows the model (TPM)architecture
sequential requires historical data for
of an RNN-based
the environmental parameters To and Ti. For each environmental
environmental prediction model for predicting the indoor temperature and humidity. parameter, the TPM requires
96 sequential data. Eachindoor
In the RNN-based row ofenvironment
the input dataset is comprised
prediction model, of thea input
TS, To,parameters
and Ti. Theare 96continuous
sequential
data are the historical data from t − 96 to t. Each t represents the number
in the time axis, and the output parameter is also continued from the inputs. An RNN model enables of TSs that present 15 min.
The t means
efficient the number
learning, to handlefor the current data
sequential time with
in thetemporal
TS. Therefore, 96 of TS
correlation presents
based 1 h. Through
on connecting the
hidden
TPM, the Ti of t + 1 can be predicted. Similarly, the humidity prediction
layers with the former ones circularly. In the proposed RNN-based model, eight hidden layers are model (HPM) requires the
historical
connecteddata withoftheHoLSTM
and Hi. Thewhich
cells, outputareofthe
therecurrent
HPM is the Hi to
units, t + 1.the historical information from
of save
The prediction
the sequence. In an models
RNN model,are built by the
there are RNN-based ML, which between
numerous connections is good incurrent
scenariosandwhere there
previously
are temporal or order dependencies through containing a memory component,
hidden layers. However, training such a network becomes difficult because of the vanishing gradient to pass the selective
knowledge
problem [65]. down
The the timecells
LSTM sequence. Figure 3 to
were proposed shows
buildthe sequential
recurrent architecture
units for the hiddenof an layers
RNN-based
in the
environmental prediction model for predicting the indoor temperature
network of RNN model. The LSTM cells are used for remembering the short-term memory that and humidity.
In the
delivers RNN-based
information to indoor environment
the result. The inputprediction
parameters model, the inputas
are presented parameters
x(n), which areare
continuous
continuousin
the
data.time
Eachaxis, and
x has the dimensions:
three output parameter TS, To,isand
alsoTicontinued
for TPM; and from the
TS, inputs.
Ho, and HiAn forRNN
HPM. model
In theenables
model,
efficient learning, to handle sequential data with temporal correlation based
the last outputs are taken to the fully-connected layer that outputs the power consumption value. on connecting hidden
layers with the former ones circularly. In the proposed RNN-based model, eight hidden layers are
connected with the LSTM cells, which are the recurrent units, to save the historical information from
the sequence. In an RNN model, there are numerous connections between current and previously
hidden layers. However, training such a network becomes difficult because of the vanishing gradient
problem [65]. The LSTM cells were proposed to build recurrent units for the hidden layers in the
network of RNN model. The LSTM cells are used for remembering the short-term memory that
delivers information to the result. The input parameters are presented as x(n), which are continuous
Sustainability 2019, 11, 997 7 of 18

data.
Sustainability x has11,
Each 2019, three dimensions:
x FOR PEER REVIEW TS, To, and Ti for TPM; and TS, Ho, and Hi for HPM. In the model,
7 of 18
the last outputs are taken to the fully-connected
Sustainability 2019, 11, x FOR PEER REVIEW layer that outputs the power consumption value.
7 of 18

Predicted Indoor Environmental Parameter y

Predicted Indoor Environmental Parameter y

Fully-Connected Layer
Input Dimension: 3
RNN Hidden Dimension: 8 Fully-Connected Layer
Input Dimension: 3
Output Dimension: 1
RNN Hidden Dimension: 8 RNN Outputs
Output Dimension: 1
RNN Outputs

LSTM LSTM LSTM LSTM

LSTM LSTM LSTM LSTM
LSTM LSTM LSTM LSTM
LSTM LSTM LSTM LSTM

x1 x2 x3 x96
x1 x2 x3 x96
Figure 3. Recurrent neural network (RNN)-based indoor environment prediction model.
Figure 3. Recurrent neural network (RNN)-based indoor environment
environment prediction
prediction model.
model.
Figure 4 presents an architecture of the proposed occupant comfort management based on the
proposed
Figureenvironment
4 presents
presents an anprediction model.
architecture
architecture of theThe
of the inputsoccupant
proposed
proposed of the environment
occupant comfort
comfort managementprediction
management model
based are
on the
collected
proposed from
proposedenvironmentan environment
environmentprediction where
predictionmodel.the occupant
TheThe
model. inputs lives.
of theof
inputs The required
environment data for
prediction
the environment the environmental
model are
prediction collected
model are
parameters
from
collected is for
from an 1environment
an environment hour,
where which is 96 the
the where sequential
occupant lives. The
occupant values.
lives.Through
requiredThedata the
for
required theprediction model,
environmental
data for the values of
parameters
environmental
temperature
is
parameters isand
for 1 h, which forishumidity
196hour, are derived
sequential
which values. for the next
Through
is 96 sequential the time. The values
prediction
values. Through model, are usedoffor
values
the prediction the PMV
temperature
model, and
values of
optimization
humidity
temperature models
are and
derived to determine
for the are
humidity a power
nextderived
time. The value.
forvalues The power
are used
the next value
time. for is used
Thethevalues
PMV and for controlling
optimization
are used an actuator
for the models
PMV and to
to affect the
determine
optimization environment.
a power
models value. The power
to determine value isvalue.
a power used for
Thecontrolling
power value an is
actuator
used forto controlling
affect the environment.
an actuator
to affect the environment.
TS t-95~t
Ti
TSt-95~t
t-95~t Environment
Environment
Hi t-95~t
Ti t-95~t Prediction Model
Environment
Environment
To
Hi t-95~t Prediction Model
Ho t-95~t
To t-95~t
Environment (t=t0=t1+1) Ti t+1 Hi t+1
Ho t-95~t
Environment Ti t+1 Hi t+1
Environment (t=t0=t1+1)
(t=t1=t0+1)
Actuator Environment (t=t1=t0+1) PMV Model
Actuator PMV Model
Power t Optimization Model
Power t Optimization Model
Figure 4. Architecture
Architecture of
of the
the proposed
proposedoccupant
occupantcomfort
comfortmanagement
managementbased
basedon
onthe
theprediction
predictionmodel.
model.
Figure 4. Architecture of the proposed occupant comfort management based on the prediction
3.2. Comfort Index Model
model.
3.2. Comfort
The PMV Index Modelin a comfort index that predicts the mean value from the votes of a large
results
group of PMV
3.2. Comfort
The people,
Index based
Model
results inon the heatindex
a comfort balance
thatof the human
predicts body.value
the mean The from
PMVthe presents
votes ofseven points
a large to
group
represent
of people, the level
based of thermal
on theinheat sensation
balance of thescale:
humanhot, warm,
body. slightly warm, neutral, slightly cool, cool,
The PMV results a comfort index that predicts theThe
meanPMV presents
value from theseven points
votes of a to represent
large group
and
the cold. The PMV
level ofbased
thermal model is
sensation used for getting
scale: of
hot, the comfort index to be used in the energy consumption
of people, on the heat balance thewarm,
human slightly
body. warm,
The PMV neutral, slightly
presents sevencool, cool,
points and cold.
to represent
optimization
The PMVofmodelmodel. The for
parameters
is sensation
used getting of the PMV-based comfort index model are the predicted
the level thermal scale: hot,the comfort
warm, index
slightly to
warm, beneutral,
used in the
slightly energy consumption
cool, cool, and cold.
environmental
optimization data
model. for the indoor
The parameters environmental parameters using TPM and HPM. Therefore, the
The PMV model is used for getting of thethe PMV-based
comfort index comfort
to be usedindexin model
the energyare the predicted
consumption
environmental
optimization model.data for Thetheparameters
indoor environmental
of the PMV-based parameters usingindex
comfort TPM model
and HPM. are Therefore,
the predictedthe
PMV value presents a predicted comfort index for the person who is in the
environmental data for the indoor environmental parameters using TPM and HPM. Therefore, the individual environment,
as
PMVwell as the
value environmental
presents datacomfort
a predicted collected to befor
index used
the in the proposed
person who is inprediction model.
the individual environment,
as well as the environmental data collected to be used in the proposed prediction model.
Sustainability 2019, 11, 997 8 of 18

PMV value presents a predicted comfort index for the person who is in the individual environment, as
well as the environmental data collected to be used in the proposed prediction model.
For calculating the PMV, Equations (1)–(8) are used. These equations refer the IOS 7730 2005
version that is mentioned in the Related Works section. The PMV formula is comprised of evaporation,
respiration, radiation, convection, and other parameters. For the overall comfort index model, the
input parameters are metabolic rate (M), effective mechanical power (W), air temperature (Ta), mean
radiant temperature (Tr), relative humidity (Rh), air velocity (Va), and clothing insulation (icl) [66–68].
The overall mathematical model is presented in Equation (1), and accordingly the parameters in
Equation (1) are further derived in Equations (2)–(5):

PMV = (0.303 × e(−0.036× M) + 0.028)

×( M − W + Evaporation + Respiration + Radiation (1)
+Convection)

Evaporation = −3.05 × 10−3 × (5733 − 6.99 × ( M − W ) − Pa) − 0.42 × ( M − W − 58.15) (2)

Respiration = −1.7 × 10−5 × M ∗ (5867 − Pa) − 0.0014 × M × (34 − Ta) (3)

Radiation = −3.96 × 10−8 × f cl × ((tcl + 273)−4 − ( Tr + 273)−4 ) (4)

Convection = − f cl × hc × (tcl − Ta) (5)

where the unit of M and W is watts per square meter (W/m2 ); the unit of Pa is pascals (Pa), and the
unit of Ta, Tr, and tcl is degrees Celsius. The parameter f cl is the clothing surface area factor that is
derived by icl, and the parameter hc is the convective heat transfer coefficient in watts per square meter
kelvin, derived from Equations (6) and (7) below:

hc1 = 2.38 × |tcl − Ta|0.25 (6)

√
hc2 = 12.1 × Va (7)

where the unit of parameter Va in Equation (7) is meters per second (m/s). The outputs hc1 and hc2
from Equations (6) and (7) are selected by the Algorithm 1 through the comparison between the results
of the equations.
For using the relative humidity in the PMV model, converting the relative humidity to saturation
vapor pressure is required. In a space, the pressure of the air is affected by humidity and temperature.
The mathematical model was given by Tetens [69], which is introduced in [70] and presented in below
with Equation (8):
Pa = Rh × (610.78 × e(17.2694×Ta/(Ta+238.3))/100 ) (8)

Algorithm 1 shows the steps of calculating the PMV based on given parameters, including M, W,
Ta, Tr, Va, Rh, and icl, using the function getPMV (M, W, Ta, Tr, Va, Rh, icl). In the first step, the value of
parameter icl is checked. Depending on whether or not the value is bigger than 0.078, the fcl is given a
different value. For the hc, according to the comparison of Equations (6) and (7), hc is given a different
value. Once the fcl and hc are confirmed, the Pa is given by Equation (8). The parameter tcl is given
by Algorithm 2. Then all required values of parameters are prepared for equations, and values are
assigned to the equations to derive the value of PMV.
Algorithm 2 shows the steps to calculate the tcl for Equations (4)–(6), which are used in Algorithm
1. For deriving the value of tcl, the input parameters fcl, M, W, Ta, Tr, Va, and icl are required. The fcl is
a result in Algorithm 1, and others are the inputs of the PMV model. The function getTcl(fcl, M, W, Ta,
Tr, Va, icl) includes an iterator to get the minimum distance between two variables, resulting in the tcl.
In this iteration the function, getTclf(tcl, fcl, ts, Ta, Tr, Va, icl) is used.
Sustainability 2019, 11, 997 9 of 18

Algorithm 1. Calculate PMV.

1: function getPMV(M, W, Ta, Tr, Va, Rh, icl)
2: if icl > 0.078 then
3: fcl = 1.05 + 0.645 * icl;
4: else
5: fcl = 1.00 + 1.290 * icl;
6: end
7: if hc1 from Equation (6) > hc2 Equation (7) then
8: hc = hc1;
9: else
10: hc = hc2;
11: end
12: Pa = Get the value using Equation (8);
13: tcl = Algorithm 2(fcl, M, W, Ta, Tr, Va, icl);
14: Evaporation = Get the value using Equation (2);
15: Respiration = Get the value using Equation (3);
16: Radiation = Get the value using Equation (4);
17: Convection = Get the value using Equation (5);
18: PMV = Get the value using Equation (1);
19: return PMV;
20: end function

Algorithm 2. Calculate tcl.

1: function getTcl(fcl, M, W, Ta, Tr, Va, icl)
2: ts = 35.7 − 0.028 * (M − W);
3: eps = 1e − 9;
4: q = {−50, 50};
5: while (|q[0] − q[1]|) > eps) {
6: temp = (q[0] * getTclf(q[1], fcl, ts, Ta, Tr, Va, icl)
7: −q[1] * getTclf (q[0], fcl, ts, Ta, Tr, Va, icl))
8: /( getTclf (q[1], fcl, ts, Ta, Tr, Va, icl)
9: −getTclf (q[0], fcl, ts, Ta, Tr, Va, icl));
10: q[0] = q[1];
11: q[1] = temp;
12: }
13: return (q[0] + q[1]) / 2;
14: end function
15: function getTclf(tcl, fcl, ts, Ta, Tr, Va, icl)
16: t1 = (tcl + 273)ˆ(−4) − (Tr + 273)ˆ(−4);
17: if hc1 from Equation (7) > hc2 Equation (8) then
18: hc = hc1;
19: else
20: hc = hc2;
21: end
22: tcle = ts − icl * (3.96 * 1e − 8 * fcl * t1 + fcl * hc * (tcl − Ta));
23: return tcl − tcle;
24: end function

To use the proposed comfort index model based on the PMV, the parameters M, W, Va, and icl
are fixed, because of the lack of training data for the environmental prediction model. The proposed
prediction models are used for resulting in the indoor temperature and humidity, which are the inputs
Sustainability 2019, 11, 997 10 of 18

of the comfort index model. Therefore, we assume other parameters for the occupant in the individual
environment. The value of M is 58, W is 0, Va is 0, and icl is 1.1. In this case, these parameters refer to a
person in a seated, quiet state, wearing long shirts and pants, when the wind cannot be detected by the
person. Based on the experiment with the given values for the input parameters of the PMV model,
the PMV value is derived correctly to compare with the American Society of Heating, Refrigerating
and Air-Conditioning Engineers (ASHRAE) Standard 55-2017 from Center for the Built Environment
(CBE) Thermal Comfort Tool [71].

3.3. Energy Consumption Optimization Model

The energy consumption model is a multi-objective function, and addresses minimizing energy
while simultaneously achieving thermal comfort. Our formula is based on the derivation of House and
Smith [22], where the cost is defined in terms of two distinct subparts. The first part is the differential
quantity of the current temperature and predicted temperature at the next time interval (t + 1), and the
second is the PMV subpart, which is also the differential quantity of the current and predicted PMVs.
In addition to these two parameters, the predicted energy consumption ( Eht+1 ) is also considerd as
given in Equation (9):

( Eht+1 )3
Z t +1 n o
f ( Eht+1 ) = α ∗ (( Tt − Tt+1 )2 + ( PMVt − PMVt+1 )2 ) dt + , α = 109 (9)
t 3

where f ( Eht+1 ) is the objective function of energy consumption optimization; α is a weighting factor,
and is taken as 10ˆ9; Tt is the generated temperature; and Tt+1 is the predicted temperature at time
interval t + 1. These change with respect to time in these two differential quantities are integrated into
the interval t and t + 1.
Equation (9) is further expanded to form Equation (10):

( Tt − Tt+1 )3 ( Tt−1 − Tt )3 ( PMVt − PMVt+1 )3 ( PMVt−1 − PMVt )3 3

f ( Eht+1 ) = α ∗ ( 3 − 3 + 3 − 3 ) + (Eht3+1 ) , α = 109 . (10)

The value of the energy consumption optimization model is generated by an integration by parts
formula, and the result shows that the cube average filter is applied to the main parts of the model, i.e.,
temperature and PMV.
Figure 5 shows the schematic of the energy consumption optimization model, using an objective
function for operating the heater to change the indoor environment, including temperature and
humidity. The TPM and HPM are adapted to provide the predicted indoor environment data for the
objective function. Through the predicted indoor temperature and humidity data, the PMV, which
is a parameter of the objective function, is derived for the future environment. The future indoor
temperature is also a parameter of the objective function. From the historical dataset, the indoor
temperature data for the time sequence t − 1 and t are required, and the PMV data for the same
time sequence with the indoor temperature are required. The optimization model outputs the power
consumption for the heater, which is a value to present the energy consumption from a time sequence
t to t + 1. The heater consumes power according to the output of the optimization model, to affect the
indoor environment. In the time t, the outdoor and indoor environments are the factors of the heater.
From these factors, the heater updates the indoor environment based on the power value, in order to
achieve the indoor environment in time t + 1.
In the implementation, we developed a Deep Neural Network (DNN)-based heater emulator
trained by the same dataset with applied to the TPM and HPM. Therefore, when the indoor and
outdoor temperature and humidity data for time t and the power value for time t + 1 are inputted into
the heater emulator, the indoor temperature and humidity for time t + 1 are outputted.
Sustainability 2019, 11, 997 11 of 18
Sustainability 2019, 11, x FOR PEER REVIEW 11 of 18

Historical Data TS: Time Sequence

To: Outdoor Temperature
(Ti)
Ti: User Desired Indoor Temperature
(Ti) Optimization Model Ho: Outdoor Humidity
(PMV) Using Objective Hi: User Desired Indoor Humidity
(PMV) Function Eh: Heater Power Consumption

Predicted Data Outdoor Environment

(Eh)
(Ti) (To, Ho)
(PMV)
Heater
Indoor Environment
(Ti, Hi)

(Ti, Hi)
Indoor Environment

Figure 5. Operation of a heater based on an optimization model using an objective function.

function.

4. Experimental Results
In the implementation, we developed a Deep Neural Network (DNN)-based heater emulator
trained
Forby the same dataset
experimenting withwith applied tooccupant
the proposed the TPM comfort
and HPM. Therefore, when
management model,the indoor
using and
energy
outdoor temperature
optimization and humidity
and environmental data for for
prediction timethe
t and the power
individual homevalue for time t +we
environment, 1 are inputted
present the
into the heater emulator, the indoor temperature and humidity for time t + 1 are outputted.
indoor temperature and humidity data comparison between the actual data and predicted results.
The TensorFlow 1.8 was used to implement the environment prediction model with the ORNL dataset.
4. Experimental
The ORNL dataset Results
is collected from a house in Compbell, United States. The dataset has 35,040 rows
and includes various
For experimenting sensing
withdata.
the The indoor occupant
proposed and outdoor temperature
comfort and humidity
management model,data,
using as well as
energy
the heater power consumption data for a room of the house were used in
optimization and environmental prediction for the individual home environment, we present the our experiment. The indoor
data,
indoor including temperature,
temperature and humidityhumidity,
data and energy consumption
comparison between the of the heater,
actual data andwaspredicted
collectedresults.
from a
great room titled as “Great RM/Room” in the dataset. The indoor environment
The TensorFlow 1.8 was used to implement the environment prediction model with the ORNL of the room is controlled
by the user,
dataset. Thewho
ORNL adjusts the heater
dataset and other
is collected fromappliances
a house intoCompbell,
achieve comfort
Unitedenvironment. We present
States. The dataset has
the prediction result of the PMV data, which is derived by the predicted temperature
35,040 rows and includes various sensing data. The indoor and outdoor temperature and humidity and humidity for
the future. Once the predicted indoor environment data is gathered from
data, as well as the heater power consumption data for a room of the house were used in our the TensorFlow-based RNN
model, then the
experiment. ThePMV indoordatadata,
is determined
including through the Java
temperature, application,
humidity, and which
energyis consumption
the implementationof the
of comfort index model. We present the environmental and PMV data
heater, was collected from a great room titled as “Great RM/Room” in the dataset. The indoorassigned to the optimization
model in orderoftothe
environment getroom
the optimal powerby
is controlled consumption
the user, whofor the heater.
adjusts theMicrosoft
heater and Office
otherExcel was used
appliances to
for deriving
achieve the environment.
comfort optimal valueWe based the proposed
present objective
the prediction resultfunction.
of the PMV Fordata,
evaluating
which isthe result by
derived of
proposed optimization
the predicted temperature model,andthe optimalfor
humidity heater
the power
future.consumption data is indoor
Once the predicted presented, and the indoor
environment data
temperature and humidity data comparisons between actual data and
is gathered from the TensorFlow-based RNN model, then the PMV data is determined through optimal results are presented.
the
Java application, which is the implementation of comfort index model. We present the environmental
4.1. Indoor Environmental Parameters Predictions
and PMV data assigned to the optimization model in order to get the optimal power consumption
Figure
for the 6 shows
heater. the data
Microsoft comparison
Office Excel wasbetween
used foractual indoor
deriving the temperature
optimal valueand predicted
based indoor
the proposed
temperature. The unit
objective function. Forof evaluating
the presentedthedata is degrees
result of proposed (ºC), and the x-axis
Celsiusoptimization model,presents the sequence
the optimal heater
numbers. The actualdata
power consumption dataisispresented,
presentedandfor the
10,512 values
indoor that showand
temperature thehumidity
indoor temperature in the
data comparisons
experimental
between actual house
data from 13 September
and optimal resultsatare
11:45 to 31 December at 23:30. The average temperature of
presented.
actual data was 21.63 ◦ C, the minimum temperature was 17.63 ◦ C, and the maximum temperature was
25.79 ◦ C. The
4.1. Indoor Environmental
predicted dataParameters
were thePredictions
results from the TPM. The predicted data are also presented for
10,512Figure
values. The average temperature of the predicted data was 21.57 ◦ C, the minimum temperature
6 shows the data comparison between actual indoor temperature and predicted indoor
was 18.03 ◦ C, The
temperature. and unit
the maximum temperature
of the presented data iswas
◦ C. The mean absolute percentage error
25.53 Celsius
degrees (ºC), and the x-axis presents the
sequence numbers. The actual data is presented for 10,512 values that show the indoor temperature
in the experimental house from 13 September at 11:45 to 31 December at 23:30. The average
temperature of actual data was 21.63 °C, the minimum temperature was 17.63 °C, and the maximum
 Sustainability 2019, 11, x FOR PEER REVIEW 12 of 18
Sustainability 2019, 11, x FOR PEER REVIEW 12 of 18

temperature was 25.79 °C. The predicted data were the results from the TPM. The predicted data are
temperature
Sustainability was
2019, 25.79 °C. The predicted data were the results from the TPM. The predicted data
11, 997 12 are
of 18
also presented for 10,512 values. The average temperature of the predicted data was 21.57 °C, the
also presented for 10,512 values. The average temperature of the predicted data was 21.57 °C, the
minimum temperature was 18.03 °C, and the maximum temperature was 25.53 °C. The mean absolute
minimum temperature was 18.03 °C, and the maximum temperature was 25.53 °C. The mean absolute
percentage
(MAPE) error
of the (MAPE) of the
predicted predicted data with actual data is 1.05%, the minimum absolute
percentage error (MAPE)data with
of the actual data
predicted dataiswith
1.05%, the minimum
actual absolute
data is 1.05%, percentage
the minimum error is
absolute
percentage
0.00%, and error
the is 0.00%,
maximum and the
absolute maximum
percentageabsolute
error is percentage
6.72%. error is 6.72%.
percentage error is 0.00%, and the maximum absolute percentage error is 6.72%.

27
27
25
25
23
[℃]

23
[℃]

21
21
19
19
17
17
11
341
681
1021
1361
1701
2041
2381
2721
3061
3401
3741
4081
4421
4761
5101
5441
5781
6121
6461
6801
7141
7481
7821
8161
8501
8841
9181
9521
9861
10201
341
681
1021
1361
1701
2041
2381
2721
3061
3401
3741
4081
4421
4761
5101
5441
5781
6121
6461
6801
7141
7481
7821
8161
8501
8841
9181
9521
9861
10201
Time Sequence
Time Sequence
Actual Indoor Temperature Predicted Indoor Temperature
Actual Indoor Temperature Predicted Indoor Temperature

Figure 6. Data comparison between the actual indoor temperature and predicted indoor
Figure6.6.Data
Data comparison between the actual indoor temperature and predicted indoor
temperature. comparison between the actual indoor temperature and predicted indoor temperature.
Figure
temperature.
Figure
Figure 77 shows
shows thethe data
data comparison
comparison between actual indoor
between actual indoor humidity
humidity and and predicted
predictedindoorindoor
FigureThe
humidity. 7 shows
unit ofthe
thedata comparison
presented data between
is actual
percentage indoor
(%), and humidity
the x-axis and predicted
presents the indoor
sequence
humidity. The unit of the presented data is percentage (%), and the x-axis presents the sequence
humidity. The unit
numbers. actualof data
the presented
is presented
presenteddataforis percentage
10,512 values(%),that
values andshow
the x-axis presents
the indoor
indoor the sequence
humidity
numbers. The actual data is for 10,512 that show the humidity ininthe
the
numbers.
experimentalThe actual
house fromdata
13 is presented
September at for
11:4510,512
to 31 values
Decemberthat
at show
23:30. the
The indoor
averagehumidity
humidity inofthe
the
experimental house from 13 September at 11:45 to 31 December at 23:30. The average humidity of the
experimental
actual data was house
53.4%,from
the 13 September
minimum at 11:45 was
humidity to 3134.44%,
Decemberandatthe
23:30. The average
maximum humidity
humidity was of the
64.49%.
actual data was 53.4%, the minimum humidity was 34.44%, and the maximum humidity was 64.49%.
actual
The data was 53.4%, the minimum humidity was 34.44%, data
and the
aremaximum humidity was 64.49%.
Thepredicted
predicteddata
data resulted
resulted from
from the
the HPM.
HPM. The The predicted
predicted data are alsopresented
also presented for10,512
for 10,512 values.
values.
The
The predicted
average data
humidity resulted
of the from the
predictedHPM.data The
waspredicted
53.21%, data
the are
minimumalso presented
humidity for
was10,512 values.
34.37%, and
The average humidity of the predicted data was 53.21%, the minimum humidity was 34.37%, and the
Themaximum
the average humidity
humidity ofwas
the predicted
64.49%. dataMAPE
The was 53.21%,
of the the minimum
predicted data humidity
with wasdata
actual 34.37%,
is and the
0.56%, the
maximum humidity was 64.49%. The MAPE of the predicted data with actual data is 0.56%, the
maximum
minimum humidity was 64.49%. The MAPE of the predicted data with actual data is 0.56%, the
minimumabsolute
absolutepercentage
percentage error
error is
is 0.00%,
0.00%, and
and the
the maximum absolute percentage
maximum absolute percentage error
error is
is 4.53%.
4.53%.
minimum absolute percentage error is 0.00%, and the maximum absolute percentage error is 4.53%.

70
70
65
65
60
60
55
55
[%]

50
[%]

50
45
45
40
40
35
35
30
30
11
341
681
1021
1361
1701
2041
2381
2721
3061
3401
3741
4081
4421
4761
5101
5441
5781
6121
6461
6801
7141
7481
7821
8161
8501
8841
9181
9521
9861
10201
341
681
1021
1361
1701
2041
2381
2721
3061
3401
3741
4081
4421
4761
5101
5441
5781
6121
6461
6801
7141
7481
7821
8161
8501
8841
9181
9521
9861
10201

Time Sequence
Time Sequence
Actual Indoor Humidity Predicted Indoor Humidity
Actual Indoor Humidity Predicted Indoor Humidity

Figure 7.
Figure 7. Data
Data comparison between actual indoor
indoor humidity
humidityand
andpredicted
predictedindoor
indoorhumidity.
humidity.
Figure 7. Data comparison between actual indoor humidity and predicted indoor humidity.
4.2.
4.2.Indoor
IndoorEnvironmental
Environmental User
User Comfort
Comfort Prediction
Prediction
4.2. Indoor Environmental User Comfort Prediction
Figure 8 shows the PMV data based on the predicted indoor temperature and humidity data
using TPM and HPM. Through the TPM, the indoor temperature data is predicted based on the TS, Ti,
and To. Using the same RNN architecture, the HPM is used for predicting the indoor humidity data,
 Sustainability 2019, 11, x FOR PEER REVIEW 13 of 18

Figure
Sustainability 8 shows
2019, the PMV data based on the predicted indoor temperature and humidity13data
11, 997 of 18
using TPM and HPM. Through the TPM, the indoor temperature data is predicted based on the TS,
Ti, and To. Using the same RNN architecture, the HPM is used for predicting the indoor humidity
based
data, on the TS,
based on Hi,
the and
TS, Ho. The PMV
Hi, and model
Ho. The PMVderives
modelthe PMV data
derives using data
the PMV theseusing
predicted
thesetemperature
predicted
and humidity data. The other parameters are fixed values for the inputs of
temperature and humidity data. The other parameters are fixed values for the inputs ofthe model. Thethelength
model.of
the
Thepresented
length ofPMV data is 10,512,
the presented PMVwhich
data presents
is 10,512,the future
which comfortthe
presents index of the
future environment
comfort index ofbased
the
on the predicted temperature and humidity. The minimum comfort index
environment based on the predicted temperature and humidity. The minimum comfort index is − 1.1, then the is
indoor
-1.1,
temperature is 18.30 ◦ C and indoor humidity is 41.64%. The maximum comfort index is 0.81, then the
then the indoor temperature is 18.30 °C and indoor humidity is 41.64%. The maximum comfort index
indoor temperature is 25.28 ◦ C and indoor humidity is 61.35%.
is 0.81, then the indoor temperature is 25.28 °C and indoor humidity is 61.35%.

1.00

0.50
[Comfort Index]

0.00

-0.50

-1.00

-1.50
1
349
697
1045
1393
1741
2089
2437
2785
3133
3481
3829
4177
4525
4873
5221
5569
5917
6265
6613
6961
7309
7657
8005
8353
8701
9049
9397
9745
10093
Time Sequence

Figure
Figure 8. Predicted
8. Predicted mean
mean vote
vote (PMV)
(PMV) data
data based
based on on predicted
predicted indoor
indoor temperature
temperature andand humidity
humidity data.
data.
4.3. Optimal Energy Consumption
4.3. Figure
Optimal9Energy
showsConsumption
power consumption data based on the energy optimization model to operate
the heater.
FigureUsing
9 showsthepower
indoor temperature,
consumption humidity,
data based onand
the PMV
energydata as the parameters
optimization model to of the input
operate the
parameters for the optimization model, the optimal power value is
heater. Using the indoor temperature, humidity, and PMV data as the parameters ofderived be used in the heater.
input
The purposefor
parameters of the
the optimization
optimal energy consumption
model, the optimalvalue
poweris value
supporting high-level
is derived be usedcomfort, as well
in the heater. Theas
minimum
purpose of energy. To compare
the optimal withconsumption
energy actual energy value
consumption data, thehigh-level
is supporting optimizedcomfort,
power consumption
as well as
minimum
data energy.
is reduced To compare
by 8.43% for 10,400with actual values
collected energyin consumption
the experiment. data,
Thethe optimized
total consumptionpowerof
consumption
actual data wasdata is reduced
414,887.06 Watt by 8.43%
(W), and thefor total
10,400 collectedpower
optimized valuesconsumption
in the experiment. The total
was 382,622.91 W.
consumption
The average ofoftheactual datadata
actual waswas
414,887.06 Watt (W), (Wh),
39.89 Watt-hour and the total
the optimized
minimum waspower
26.75consumption
Wh, and the
was 382,622.91
maximum W. The
was 65.41 Wh.average of the of
The average actual data was 39.89
the optimized power Watt-hour (Wh),isthe
consumption minimum
36.79 Wh, thewas 26.75
minimum
isWh,
8.20and
Wh,the
andmaximum was 65.41
the maximum Wh. The
is 127.95 Wh.average of the optimized power consumption is 36.79 Wh,
the minimum is 8.20 Wh, and the maximum is 127.95 Wh.
Sustainability 2019, 11, 997 14 of 18
Sustainability 2019,
Sustainability 2019, 11,
11, xx FOR
FOR PEER
PEER REVIEW
REVIEW 14 of
14 of 18
18

140
140
120
120
100
100
80
[Wh]

80
[Wh]

60
60
40
40
20
20
00
11
337
673
1009
1345
1681
2017
2353
2689
3025
3361
3697
4033
4369
4705
5041
5377
5713
6049
6385
6721
7057
7393
7729
8065
8401
8737
9073
9409
9745
10081
337
673
1009
1345
1681
2017
2353
2689
3025
3361
3697
4033
4369
4705
5041
5377
5713
6049
6385
6721
7057
7393
7729
8065
8401
8737
9073
9409
9745
10081
Time Sequence
Time Sequence

Actual Power
Actual Power Consumption
Consumption Optimized Power
Optimized Power Consumption
Consumption

Figure 9.
Figure
Figure 9. Optimal power
9. power consumption
consumption data
consumption data for
data for operating
foroperating heater.
operatingheater.
heater.

4.4.
4.4.Indoor
IndoorEnvironment
EnvironmentUsing
Using Optimal
Optimal Power
Power Consumption
Consumption
4.4. Indoor Environment Using Optimal Power Consumption
Figure
Figure10
Figure 10shows
10 showsthe
shows the data
the data comparison
data between
comparison between
comparison the actual
between the
the actualindoor
actual indoortemperature
indoor temperatureand
temperature andoptimal
and optimalindoor
optimal indoor
indoor
temperature. ◦
temperature.The
temperature. Theaverage
The averagetemperature
average temperatureofof
temperature the
of actual
the
the data
actual
actual was
data
data was
was21.60
21.60
21.60 C,°C,
thethe
°C, minimum
the minimum
minimumtemperature was
temperature
temperature
17.67 ◦ C, and the maximum temperature was 25.79 ◦ C. The average temperature of the optimal data
was 17.67 °C, and the maximum temperature was 25.79 °C. The average temperature
was 17.67 °C, and the maximum temperature was 25.79 °C. The average temperature of the optimal of the optimal
◦ C, the minimum temperature was 18.26 ◦ C, and the maximum temperature was 25.51 ◦ C.
was
data21.51
data was 21.51
was 21.51 °C,
°C, the
the minimum
minimum temperature
temperature was
was 18.26
18.26 °C, and the
°C, and the maximum
maximum temperature
temperature was
was 25.51
25.51
The
°C. The MAPE of the optimal data with the actual data is 0.42%, the minimum absolute percentageis
°C. MAPE
The of
MAPE the optimal
of the data
optimal with
data the
withactual
the data
actualis 0.42%,
data is the
0.42%,minimum
the absolute
minimum percentage
absolute error
percentage
0.00%,
error isand
error is the maximum
0.00%,
0.00%, and the
and absoluteabsolute
the maximum
maximum percentage
absolute error is 4.27%.
percentage
percentage error is
error is 4.27%.
4.27%.

27
27
25
25
23
23
[℃]
[℃]

21
21
19
19
17
17
11
373
745
1117
1489
1861
2233
2605
2977
3349
3721
4093
4465
4837
5209
5581
5953
6325
6697
7069
7441
7813
8185
8557
8929
9301
9673
10045
373
745
1117
1489
1861
2233
2605
2977
3349
3721
4093
4465
4837
5209
5581
5953
6325
6697
7069
7441
7813
8185
8557
8929
9301
9673
10045

Time Sequence
Time Sequence

Actual Indoor
Actual Indoor Temperature
Temperature Optimal Indoor
Optimal Indoor Temperature
Temperature

Figure10.
Figure
Figure 10.Data
10. Data comparison
Data comparison between
comparison between actual
actual indoor
indoor temperature
temperature and
temperature and optimal
and optimal indoor
optimalindoor temperature.
indoortemperature.
temperature.

Figure11
Figure
Figure 11shows
11 showsthe
shows the data
the data comparison
data comparison between
comparison between the
the actual
actual indoor
actual indoor humidity
indoor humidity
humidityand and optimal
andoptimal indoor
optimalindoor
indoor
humidity. The
humidity.The
humidity. average
Theaverage humidity
average humidity
humidity of of the actual data was
of the actual data was 53.36%,
was 53.36%, the
53.36%, the minimum
theminimum humidity
minimumhumidity was
humiditywas 34.44%,
was34.44%,
34.44%,
andthe
and
and themaximum
the maximumhumidity
maximum humidity was
humidity was 64.72%.
was 64.72%. The
64.72%. The average
average humidity
average humidity
humidity ofof the
of the optimal
theoptimal data
optimaldata was
datawas 53.36%,
was53.36%, the
53.36%,the
the
minimum humidity
minimumhumidity
minimum was
humiditywas 23.66%,
was23.66%, and
23.66%, and the maximum humidity
and the maximum humidity was was 64.45%.
was 64.45%.
64.45%. TheThe MAPE
TheMAPE of the
MAPEofofthe optimal
theoptimal
optimal
datawith
data
data withactual
with actualdata
actual dataisis
data is9.99%,
9.99%, the
9.99%, the minimum
the minimum absolute
minimum absolute percentage
absolute percentage error
percentage error
errorisis 0.04%,
is 0.04%, and
0.04%,and the
andthe maximum
themaximum
maximum
absolutepercentage
absolute
absolute percentageerror
percentage errorisis
error is46.09%.
46.09%.
46.09%.
Sustainability 2019, 11, 997 15 of 18
Sustainability 2019, 11, x FOR PEER REVIEW 15 of 18

70

60

50
[%]

40

30

20
1
373
745
1117
1489
1861
2233
2605
2977
3349
3721
4093
4465
4837
5209
5581
5953
6325
6697
7069
7441
7813
8185
8557
8929
9301
9673
10045
Time Sequence

Actual Indoor Humidity Optimal Indoor Humidity

Figure11.
Figure 11. Data
Data comparison
comparison between actual indoor
indoor humidity
humidity and
and optimal
optimalindoor
indoorhumidity.
humidity.

5.5.Conclusions
Conclusions
In
Inthis
thispaper,
paper, wewe propose
propose an improved
an improved energy optimization
energy model
optimization to recommend
model to recommend optimal power
optimal
consumption for operating heaters, which is necessary to change the indoor
power consumption for operating heaters, which is necessary to change the indoor environmental environmental factors
like temperature
factors and humidity.
like temperature We used indoor
and humidity. We usedtemperature and humidityand
indoor temperature as input parameters
humidity as inputfor
the optimization model. We also used RNN and LSTM for prediction
parameters for the optimization model. We also used RNN and LSTM for prediction models to models to forecast energy
consumption.
forecast energy The results of theThe
consumption. prediction model
results of suggest that
the prediction the PMV
model suggest values leadPMV
that the to the derivation
values lead
oftothe
thecomfort
derivationindexof values, whichindex
the comfort can be applied
values, to thecan
which proposed optimization
be applied model. optimization
to the proposed In the present
experiment,
model. In the wepresent
have exploited
experiment,MAPE to predict
we have indoor
exploited temperature
MAPE to predictvalues
indoorand indoor humidity
temperature values
values that have
and indoor 1.05%values
humidity and 0.56% differences
that have 1.05% with the actual
and 0.56% data, respectively.
differences Based
with the actual on respectively.
data, the predicted
Based
data on relation
with the predicted
to thedata
PMVwith relation
data, to the PMV
the optimized data,
power the optimized
consumption forpower
10,400consumption
records result forin
an10,400
8.43%records resultcompared
reduction in an 8.43%toreduction
the actualcompared to the actual
data. Similarly, data.
on the Similarly,
basis of theon the basis of
optimized the
power
optimized power
consumption, consumption,
the indoor the indoor
temperature datatemperature data results
results in a 0.42% in a 0.42%
and indoor and indoor
humidity humidity
data results in a
data results
9.99% in a with
difference 9.99%thedifference withConsequently,
actual data. the actual data.theConsequently,
experimental the experimental
results dictate that results dictate
the proposed
that the proposed
optimization scheme savesoptimization
energy, asscheme saves energy,
well as providing as well environment,
a comfortable as providingusing a comfortable
the optimal
environment,
power consumption usingtothe optimal
operate thepower
heaterconsumption
and thus keeptothe operate the heater
user-desired and thusand
temperature keep the user-in
humidity
desired
the indoor temperature
environment. and humidity in the indoor environment.

Author Contributions:W.J.,
AuthorContributions: W.J.,I.U.,
I.U.,S.A.
S.A.and
andD.H.K.
D.H.K.designed
designed the
the overall
overall system.
system. W.J. implemented theoverall
implemented the overall
system
systemand
andperformed
performed experiments. W.J. wrotethis
experiments. W.J.wrote thispaper.
paper.

Funding:This
Funding: Thisresearch
researchwas
wasfunded
funded byby Korea
Korea government
government (MSIT)
(MSIT) grant number 2018-0-01456.
grant number 2018-0-01456.
Acknowledgments: This work was supported by Institute for Information & communications Technology
Acknowledgments: This work was supported by Institute for Information & communications Technology
Promotion (IITP) grant funded by the Korea government (MSIT) (No.2018-0-01456, AutoMaTa: Autonomous
Promotion (IITP)
Management grant funded
framework based onby artificial
the Koreaintelligent
government (MSIT) (No.2018-0-01456,
Technology for adaptive andAutoMaTa:
disposable Autonomous
IoT), and this
Management
research framework
was supported based
by the on (Ministry
MSIT artificial intelligent Technology
of Science and for adaptive
ICT), Korea, under theand
ITRCdisposable IoT),
(Information and this
Technology
Research
research Center) support by
was supported program (IITP-2017-2016-0-00313)
the MSIT (Ministry of Science supervised by the under
and ICT), Korea, IITP (Institute
the ITRC for(Information
Information
&Technology
communications Technology
Research Promotion).
Center) support programAny(IITP-2017-2016-0-00313)
correspondence related supervised
to this paper
byshould be(Institute
the IITP addressedforto
Dohyeun Kim.
Information & communications Technology Promotion). Any correspondence related to this paper should be
addressed
Conflicts ofto Dohyeun
Interest: Kim.
The authors declare no conflict of interest.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.References
Rupp, R.F.; Vásquez, N.G.; Lamberts, R. A review of human thermal comfort in the built environment.
1. Energy
Rupp, Build. 2015, 105,N.G.;
R.F.; Vásquez, 178–205. [CrossRef]
Lamberts, R. A review of human thermal comfort in the built environment.
2. Yang, L.; Yan, H.; Lam, J.C. Thermal
Energy Build. 2015, 105, 178–205. comfort and building energy consumption implications—A review.
2. Appl.
Yang,Energy H.;115,
2014,
L.; Yan, Lam,164–173. [CrossRef]
J.C. Thermal comfort and building energy consumption implications—A review.
Appl. Energy 2014, 115, 164–173.
Sustainability 2019, 11, 997 16 of 18

3. Chu, Y.; Xu, P.; Yang, Z.; Li, W. Existing building retrofitting for Indoor PM2.5 Concentration Control on
Smog Days: Case study in China. Procedia Eng. 2017, 205, 3222–3227. [CrossRef]
4. UN. World Urbanization Prospects: The 2014 Revision-Highlights; UN: New York, NY, USA, 2014.
5. MacNaughton, P. Green Buildings and Health. 2016. Available online: https://dash.harvard.edu/handle/1/
32644538 (accessed on 1 December 2018).
6. Al Horr, Y.; Arif, M.; Kaushik, A.; Mazroei, A.; Katafygiotou, M.; Elsarrag, E. Occupant productivity and
office indoor environment quality: A review of the literature. Build. Environ. 2016, 105, 369–389. [CrossRef]
7. Haynes, B.; Suckley, L.; Nunnington, N. Workplace productivity and office type: An evaluation of office
occupier differences based on age and gender. J. Corp Real Estate 2017, 19, 111–138. [CrossRef]
8. Geng, Y.; Ji, W.; Lin, B.; Zhu, Y. The impact of thermal environment on occupant IEQ perception and
productivity. Build. Environ. 2017, 121, 158–167. [CrossRef]
9. Shaikh, P.H.; Nor, N.B.M.; Nallagownden, P.; Elamvazuthi, I.; Ibrahim, T. A review on optimized control
systems for building energy and comfort management of smart sustainable buildings. Renew. Sustain.
Energy Rev. 2014, 34, 409–429. [CrossRef]
10. Wang, Z.; Yang, R.; Wang, L. Multi-agent intelligent controller design for smart and sustainable buildings.
In Proceedings of the 2010 IEEE International Systems Conference, San Diego, CA, USA, 5–8 April 2010.
11. Yang, R.; Wang, L. Optimal control strategy for HVAC system in building energy management.
In Proceedings of the PES T&D 2012, Orlando, FL, USA, 7–10 May 2012; pp. 1–8.
12. Dounis, A.; Caraiscos, C. Advanced control systems engineering for energy and comfort management in a
building environment—A review. Renew. Sustain. Energy Rev. 2009, 13, 1246–1261. [CrossRef]
13. Laftchiev, E.; Nikovski, D. An IoT system to estimate personal thermal comfort. In Proceedings of the 2016
IEEE 3rd World Forum on Internet of Things (WF-IoT), Reston, VA, USA, 12–14 December 2016; pp. 672–677.
14. Farhan, A.A.; Pattipati, K.; Wang, B.; Luh, P. Predicting individual thermal comfort using machine
learning algorithms. In Proceedings of the 2015 IEEE International Conference on Automation Science and
Engineering (CASE), Gothenburg, Sweden, 24–28 August 2015; pp. 708–713.
15. Afram, A.; Janabi-Sharifi, F.; Fung, A.S.; Raahemifar, K. Artificial neural network (ANN) based model
predictive control (MPC) and optimization of HVAC systems: A state of the art review and case study of a
residential HVAC system. Energy Build. 2017, 141, 96–113. [CrossRef]
16. Zupančič, D.; Luštrek, M.; Gams, M. Trade-off between Energy Consumption and Comfort Experience in
Smart Buildings. Inf. Technol. Control 2015, 44, 420–432. [CrossRef]
17. Akkaya, K.; Guvenc, I.; Aygun, R.; Pala, N.; Kadri, A. IoT-based occupancy monitoring techniques for
energy-efficient smart buildings. In Proceedings of the 2015 IEEE Wireless Communications and Networking
Conference Workshops (WCNCW), New Orleans, LA, USA, 9–12 March 2015; pp. 58–63.
18. Cannistraro, G.; Cannistraro, M.; Trovato, G. Island “Smart Energy” for Eco- Sustainable Energy—A Case
Study “Favignana Island”. Int. J. Heat Techol. 2017, 35, S87–S95. [CrossRef]
19. Suryadevara, N.K.; Mukhopadhyay, S.C.; Kelly, S.D.T.; Gill, S.P.S. WSN-Based Smart Sensors and Actuator
for Power Management in Intelligent Buildings. IEEE/ASME Trans. Mechatron. 2015, 20, 564–571. [CrossRef]
20. Cannistraro, M.; Lorenzini, E. The Applications of the New Techologies E-Sensing in Hospitals. Int. J.
Heat Techol. 2016, 34, 551–557. [CrossRef]
21. Wang, J.; Wang, J. Forecasting energy market indices with recurrent neural networks: Case study of crude oil
price fluctuations. Energy 2016, 102, 365–374. [CrossRef]
22. Long-term energy/environment data for ORNL Research House #3. Available online: https://openei.org/
datasets/dataset/ornl-research-house-3 (accessed on 1 July 2018).
23. Kim, J.; Schiavon, S.; Brager, G. Personal comfort models—A new paradigm in thermal comfort for
occupant-centric environmental control. Build. Environ. 2018, 132, 114–124. [CrossRef]
24. Mossolly, M.; Ghali, K.; Ghaddar, N. Optimal control strategy for a multi-zone air conditioning system using
a genetic algorithm. Energy 2009, 34, 58–66. [CrossRef]
25. Peeters, L.; De Dear, R.; Hensen, J.; D’Haeseleer, W. Thermal comfort in residential buildings: Comfort values
and scales for building energy simulation. Appl. Energy 2009, 86, 772–780. [CrossRef]
26. Wang, Z.; Wang, L.; Dounis, A.I.; Yang, R. Multi-agent control system with information fusion based comfort
model for smart buildings. Appl. Energy 2012, 99, 247–254. [CrossRef]
27. Molina-Solana, M.; Ros, M.; Ruiz, M.D.; Gómez-Romero, J.; Martin-Bautista, M. Data science for building
energy management: A review. Renew. Sustain. Energy Rev. 2017, 70, 598–609. [CrossRef]
Sustainability 2019, 11, 997 17 of 18

28. Široký, J.; Oldewurtel, F.; Cigler, J.; Prívara, S. Experimental analysis of model predictive control for an
energy efficient building heating system. Appl. Energy 2011, 88, 3079–3087. [CrossRef]
29. Dounis, A.; Santamouris, M.; Lefas, C. Building visual comfort control with fuzzy reasoning.
Energy Convers. Manag. 1993, 34, 17–28. [CrossRef]
30. Jabłoński, K.; Grychowski, T. Fuzzy inference system for the assessment of indoor environmental quality in
a room. Indoor Built Environ. 2017, 27, 1415–1430. [CrossRef]
31. Aste, N.; Manfren, M.; Marenzi, G. Building Automation and Control Systems and performance optimization:
A framework for analysis. Renew. Sustain. Energy Rev. 2017, 75, 313–330. [CrossRef]
32. Manic, M.; Wijayasekara, D.; Amarasinghe, K.; Rodriguez-Andina, J.J. Building Energy Management Systems:
The Age of Intelligent and Adaptive Buildings. IEEE Ind. Electron. Mag. 2016, 10, 25–39. [CrossRef]
33. Shaikh, P.H.; Nor, N.B.M.; Nallagownden, P.; Elamvazuthi, I. Stochastic optimized intelligent controller for
smart energy efficient buildings. Sustain. Cities Soc. 2014, 13, 41–45. [CrossRef]
34. Hall, D.; Llinas, J. An introduction to multisensor data fusion. Proc. IEEE 1997, 85, 6–23. [CrossRef]
35. Dounis, A.I.; Caraiscos, C. Intelligent coordinator of fuzzy controller-agents for indoor environment control
in buildings using 3-D fuzzy comfort set. In Proceedings of the 2007 IEEE International Fuzzy Systems
Conference, London, UK, 23–26 July 2007.
36. Karaboga, D. An Idea Based on Honey Bee Swarm for Numerical Optimization; Technical Report-tr06; Computer
Engineering Department, Engineering Faculty, Erciyes University: Talas, Turkey, 2005; Volume 200.
37. Fong, K.; Hanby, V.; Chow, T. System optimization for HVAC energy management using the robust
evolutionary algorithm. Appl. Therm. Eng. 2009, 29, 2327–2334. [CrossRef]
38. Wright, J.A.; Loosemore, H.A.; Farmani, R. Optimization of building thermal design and control by
multi-criterion genetic algorithm. Energy Build. 2002, 34, 959–972. [CrossRef]
39. Soares, N.; Bastos, J.; Pereira, L.D.; Soares, A.; Amaral, A.; Asadi, E.; Rodrigues, E.; Lamas, F.; Monteiro, H.;
Lopes, M.; et al. A review on current advances in the energy and environmental performance of buildings
towards a more sustainable built environment. Renew. Sustain. Energy Rev. 2017, 77, 845–860. [CrossRef]
40. Serale, G.; Fiorentini, M.; Capozzoli, A.; Bernardini, D.; Bemporad, A. Model Predictive Control (MPC)
for Enhancing Building and HVAC System Energy Efficiency: Problem Formulation, Applications and
Opportunities. Energies 2018, 11, 631. [CrossRef]
41. Huang, W.; Lam, H. Using genetic algorithms to optimize controller parameters for HVAC systems.
Energy Build. 1997, 26, 277–282. [CrossRef]
42. Delgarm, N.; Sajadi, B.; Kowsary, F.; Delgarm, S. Multi-objective optimization of the building energy
performance: A simulation-based approach by means of particle swarm optimization (PSO). Appl. Energy
2016, 170, 293–303. [CrossRef]
43. Wang, Z.; Yang, R.; Wang, L. Multi-agent control system with intelligent optimization for smart and
energy-efficient buildings. In Proceedings of the IECON 2010—36th Annual Conference on IEEE Industrial
Electronics Society, Glendale, AZ, USA, 7–10 November 2010.
44. Ullah, I.; Kim, D. An Improved Optimization Function for Maximizing User Comfort with Minimum Energy
Consumption in Smart Homes. Energies 2017, 10, 1818. [CrossRef]
45. Cetin, K.; Tabares-Velasco, P.; Novoselac, A. Appliance daily energy use in new residential buildings: Use
profiles and variation in time-of-use. Energy Build. 2014, 84, 716–726. [CrossRef]
46. Ullah, I.; Ahmad, R.; Kim, D. A Prediction Mechanism of Energy Consumption in Residential Buildings
Using Hidden Markov Model. Energies 2018, 11, 358. [CrossRef]
47. Tripathy, M. Power transformer differential protection using neural network Principal Component Analysis
and Radial Basis Function Neural Network. Simul. Model. Pract. Theory 2010, 18, 600–611. [CrossRef]
48. Niu, H.; Wang, J. Financial time series prediction by a random data-time effective RBF neural network.
Soft Comput. 2013, 18, 497–508. [CrossRef]
49. Sermpinis, G.; Stasinakis, C.; Dunis, C. Stochastic and genetic neural network combinations in trading and
hybrid time-varying leverage effects. J. Int. Financ. Mark. Inst. Money 2014, 30, 21–54. [CrossRef]
50. Schmidhuber, J. Deep learning in neural networks: An overview. Neural Netw. 2015, 61, 85–117. [CrossRef]
51. Mudassar, L.; Byun, Y. Customer Prediction on Parking Logs Using Recurrent Neural Network.
In International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed
Computing; Springer: Cham, Switzerland, 2018.
Sustainability 2019, 11, 997 18 of 18

52. Jain, V.; Seung, S. Natural image denoising with convolutional networks. In Proceedings of the
Twenty-Second Annual Conference on Neural Information Processing Systems, Vancouver, BC, Canada,
8–11 December 2008.
53. Fernández, S.; Graves, A.; Schmidhuber, J. Sequence labelling in structured domains with hierarchical
recurrent neural networks. In Proceedings of the 20th International Joint Conference on Artificial Intelligence,
IJCAI 2007, Hyderabad, India, 6–12 January 2007.
54. Wu, Z.; Ma, X.; Yu, H.; Wang, S.; Wang, Y. Spatiotemporal Recurrent Convolutional Networks for Traffic
Prediction in Transportation Networks. Sensors 2017, 17, 1501.
55. Cheng, L.; Zang, H.; Ding, T.; Sun, R.; Wang, M.; Wei, Z.; Sun, G. Ensemble Recurrent Neural Network Based
Probabilistic Wind Speed Forecasting Approach. Energies 2018, 11, 1958. [CrossRef]
56. Chemali, E.; Kollmeyer, P.J.; Preindl, M.; Ahmed, R.; Emadi, A.; Kollmeyer, P. Long Short-Term Memory
Networks for Accurate State-of-Charge Estimation of Li-ion Batteries. IEEE Trans. Ind. Electron. 2018, 65,
6730–6739. [CrossRef]
57. Tam, Y.-C.; Shi, Y.; Chen, H.; Hwang, M.-Y. RNN-based labeled data generation for spoken language
understanding. In Proceedings of the Sixteenth Annual Conference of the International Speech
Communication Association, Dresden, Germany, 6–10 September 2015.
58. Zhao, Z.; Chen, W.; Wu, X.; Chen, P.C.Y.; Liu, J. LSTM network: A deep learning approach for short-term
traffic forecast. IET Intell. Transp. Syst. 2017, 11, 68–75. [CrossRef]
59. Fanger, P.O. Thermal Comfort: Analysis and Applications in Environmental Engineering; Danish Technical Press:
Copenhagen, Denmark, 1970.
60. Olesen, B.W.; Parsons, K.C. Introduction to thermal comfort standards and to the proposed new version of
EN ISO 7730. Energy Build. 2002, 34, 537–548. [CrossRef]
61. ISO EN7730:2005. Ergonomics of the Thermal Environment-Analytical Determination and Interpretation of Thermal
Comfort Using Calculation of the PMV and PPD Indices and Local Thermal Comfort Criteria; ISO: Geneva,
Switzerland, 2005.
62. Chen, X.; Wang, Q.; Srebric, J. A data-driven state-space model of indoor thermal sensation using occupant
feedback for low-energy buildings. Energy Build. 2015, 91, 187–198. [CrossRef]
63. Ghahramani, A.; Castro, G.; Becerik-Gerber, B.; Yu, X. Infrared thermography of human face for monitoring
thermoregulation performance and estimating personal thermal comfort. Build. Environ. 2016, 109, 1–11.
[CrossRef]
64. Ari, S.; Wilcoxen, P.; Khalifa, H.E.; Dannenhoffer, J.F.; Işik, C. A practical approach to individual thermal
comfort and energy optimization problem. In Proceedings of the NAFIPS 2008—2008 Annual Meeting of the
North American Fuzzy Information Processing Society, New York, NY, USA, 19–22 May 2008.
65. Kawakami, K. Supervised Sequence Labelling with Recurrent Neural Networks. Ph.D. Thesis, Technical
University of Munich, Munich, Germany, 2008.
66. Cannistraro, M.; Cannistraro, G.; Restivo, R. The Local Media Radiant Temperature for the Calcuilation of
Comfort in Areas Characterized by Radiant Surfaces. Int. J. Heat Techol. 2015, 33, 115–122. [CrossRef]
67. Cannistraro, M.; Cannistraro, G.; Restivo, R. Some Observations on the Influence on Exchanges Radiative.
Int. J. Heat Techol. 2015, 33, 79–84. [CrossRef]
68. Cannistraro, M.; Cannistraro, G.; Restivo, R. Smart Controll of Air Climatization System in Function on the
Values of the Mean Local Radiant Temperature. Smart Sci. 2015, 3. [CrossRef]
69. Tetens, O. Uber einige meteorologische Begriffe. Z. Geophys. 1930, 6, 297–309.
70. Snyder, R.L. Humidity Conversion; Biometeorology Program, University of California: Davis, CA, USA, 2005.
71. CBE Thermal Comfort Tool. Available online: http://comfort.cbe.berkeley.edu (accessed on 12
December 2018).

© 2019 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/).