Estimation of Air-Cooling Devices Run Time Via Fuzzy Logic and Adaptive Neuro-Fuzzy Inference System

Estimation of Air-Cooling Devices Run Time Via Fuzzy Logic and
Adaptive Neuro-Fuzzy Inference System
Dina Husham Alatraqchi1, Laith A. Mohammed2
1 PG Student, Dept. of Computer Engineering, College of Technical Engineering, Mosul, Iraq
2 Dr, Dept. of Computer Engineering, College of Technical Engineering, Mosul, Iraq
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Inthis paper, fuzzy logic controller andadaptive
neuro-fuzzy inference system (ANFIS) methods were applied to
develop a run time control system for air-cooling devices. The
system uses the current temperature and door state of the
room as input variables and predicts the optimum runtime for
the device. For the fuzzy logic controller, three different
membership functions were assessed and their performance
was evaluated. The triangular membership function displayed
superior performance for the current case. The ANFIS model
was developed and validated via various validation
parameters to ensure it has the abilityto estimate theruntime
accurately. The obtained ANFIS model showed significant
validation parameters for both the training and test set. Also,
the ANFIS model was superior to the fuzzy logic controller in
terms of determining the optimum run time. Thus, the ANFIS
modeling approach can be used as an efficient and accurate
method to develop systems for controlling the run time of air
cooling devices.
Key Words: Fuzzy logic, ANFIS, Embedded system,
Microcontroller, Fuzzy controller.
For many years, the fuzzy logic controller has been an
important and popular method [1]. Due to the imprecise
nature of computer-assisted control issue solutions, the
fuzzy logic controller was created. Fuzzy logic controller
deals with data and processes it in a manner similar to
human thinking [2]. The fuzzy logic implies human-like
reasoning for determining the optimum solution. Unlike
classic logic systems where the values are considered only
exact (i.e. true or false), fuzzy systems allow vague
representation via fuzzy sets of the input values [3]. The
inference in fuzzy systems uses a set of pre-defined IF-THEN
rules to decide the output value from the inputs. An example
of a fuzzy rule would be; IF the Temperature is COLD THEN
Run Time is SHORT [4]. The output of a fuzzy system is
defuzzified to a crisp value that can be used in real-world
applications. The main advantage of using fuzzy systems is
that their structure is relatively simple and intuitive for
humans. Also, the system can be easily adjusted and
modified as required [5, 6]. The Adaptive neuro-fuzzy
inference system (ANFIS) is a hybrid learning system that
combines the fuzzy logic systems and neural network
characteristics [7]. The ANFIS modeling uses training data
combined with a set of fuzzy logic rules to produce a
machine learning model that can be used for making
predictions of the output variable from the values of the
input variables [8]. Due to the ability of neural networks to
adjust and learn the data, the ANFIS approach can provide
more powerful predictions compared to mere fuzzy logic
systems [9].
Various studies have been reported in the literature that
involves using ANFIS and fuzzy logic systems in air-cooling
devices-related applications. For instance, Soyguder et al.
[10] developed an expert system that includes ANFIS and
fuzzy logic optimization to control heating, ventilation and
air-cooling (HVAC) systems. Their system mainly focused on
controlling the humidity and the temperature of the HVAC.
The obtained ANFIS models were validated and showed low
error in terms of estimating the required parameters for
controlling the system.
In another study, Al-Jarrah et al.[11] developed an algorithm
via ANFIS modeling that focused on controlling air-cooling
systems at different pressure values. The built ANFIS model
was evaluated and assessed using experimental test data.
The predictions of the model were compared to the real
values and the computed error parameters indicated a
reliable and predictive model for managing the performance
of air-conditioners at different pressure values.
In this study, we use the fuzzy logic system and ANFIS
modeling methods to develop a system that can predict the
optimum run time for air-conditioners using the current
room temperature and the door state of the chamber as the
input variables. The current temperature of the room is a
common factor to consider when determining the optimum
running time. Also, considering that rooms with opened door
require a longer period of air cooling due to the faster heat
transfer, the door state was added as another factor for
determining the run time.
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 10 Issue: 03 | Mar 2023 www.irjet.net p-ISSN: 2395-0072
© 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 291
2. METHOD
Two different approaches were used to build a run time
prediction system for air-cooling devices, namely the fuzzy
logic system and the ANFIS. The dataset used composed of
100 records of temperature anddoor stateas input variables
1. INTRODUCTION

2.1 Fuzzy Logic:
The fuzzy logic systems work-flow can be broadly divided
intothree distinctsteps, namely,thefuzzificationoftheinput,
the application of the IF-THEN set of rules and the
defuzzification step [3]. The steps of the fuzzy system are
illustrated in Figure 1. In the upcoming sections, a brief
description of each step is provided alongside the procedure
applied for the current case.
Fig -1: Work-flow diagram of the fuzzy logic system.
2.1.1 Input Fuzzification
Initially, a set of linguistic variables is defined based on
(1)
(2)
Table-1: The values of the input temperature linguistic
variables. The a, b, and c values are used in the triangular
membership function as shown in Figure 3.
In this study, three different membership functions were
used and tested on the data, namely, the triangular,
trapezoidal and Gaussian membership functions, which are
widely used in different fuzzy logic applications [19-21]. The
shape of each membership function is shown in Figure 2.
Each membership function was assessed by making
predictions of the output variable (the run time) and
compared with the experimental optimum value. The mean
absolute error (MAE) parameter was computed for each one
and compared with one another. The lower the value of the
MAE parameter the more accurate and reliable the
membership function is considered to be [22].
35
30
25
Very Hot
30
25
20
Hot
15 20 25
Warm
Cold 10 15 20
Very Cold 5 10 15
b c
Linguistic Variable a
and the optimum run time of air conditioner as the output
variable. The implementation of the software was carried
out via the Python programming language, using mainly
the NumPy package for array-based and numerical
operations [12]. Also, the Matlab package was used for
generating the plots of the results [13]. In the upcoming
sections, a brief description o-f each method will be
provided alongside the procedures used in developing the
systems.
the current application, those linguistic variables are used in
the IF-THEN rules evaluation [14]. For example, for the
temperature input variable, five different linguistic variables
are defined; Very Cold, Cold, Warm, Hot and Very Hot. Based
on the numerical value of the input temperature, each of
those linguistic variables will be assigned a fuzzy value
(ranging from 0 to 1) [15]. For this purpose, a membership
function is used. There are various types of membership
functions with different shapes such as triangular and
trapezoidal membership functions [16]. The membership
function converts the input value to a fuzzy value via a
specific mathematical equation, for example, the equation of
the triangular membership function is depicted in (1) and
(2). Three values representing the upper, lower and medium
values of each linguistic variable are also taken
intoconsideration [17, 18]. Table 1 shows the linguistic
variables of the temperature input variable as well as the a,
b and c values required by the triangular membership
function equation for computing the fuzzy value.
Fig -2: The shapes of the three membership functions used
in this study.

2.1.3. Output defuzzification
Following the evaluation of the rules using the IF-THEN
statements, the output is a strength of each linguistic output
variable, which is a fuzzy value. The last step is the
defuzzification of the obtained values to get a crisp output
value [25]. There are various methods for performing
defuzzification. In the current system, the weighted average
method is adopted, which is simple and effective. The
equation used for computing the crisp value using this
method is shown in (3) [26].
(3)
2.1.2. IF-THEN rules
After the input fuzzification stage which determines the
degree of membership of each input linguistic variable, the
next stage involves the evaluation of rules using IF-THEN
statements [5]. For example, IF Temperature is COLD AND
Door State is OPEN THEN Run Time is LONG. To assess the
strength of these rules, the degree of membership obtained
from the input fuzzification is used. The set of rules used in
the current fuzzy system is shown in Table 2.
Table-2: The set of rules implemented in the fuzzy logic
system
Preparation of Manuscript
Temp/Door State Open
Half Open Closed
Very Cold Very Long Very Long Long
Cold Very Long Long Long
Warm
Long Medium Medium
Hot
Medium
short
Very Short
Very Hot
Very Short
Very Short
Very Short
Where is the strength of each rule and Xi is
the maximum value used in the triangular membership
function. This method yields results that are similar to the
commonly used center of area methods,however, it demands
less computations and is easier to implement [25, 26].
2.2. Adaptive Neuro-Fuzzy Inference System(ANFIS)
The ANFIS method combines the fuzzy logic system and
the neural network approaches to provide a more efficient
prediction system [8]. In particular, ituses thecharacteristics
of the fuzzy logic including the input fuzzification via a
membership function and the IF-THEN rules evaluation as
well as the learning abilities of neural networks. This makes
the system more accurate and robust with higher efficiency in
terms of learning the data and making better predictions
[27].
2.2.1. Structure and layers
Considering that ANFIS is a neural network, it is
composed of layers and nodes connected by edges that
represent weights. The input variables are transferred and
modified as they pass through the layers of the network to
produce the output value. The weights are updated during
the training process of the network to adjust the data [28].
The general structure of the ANFIS network is shown in
Figure 3.
There are different methods to compute the strength of each
rule. The method applied in the current system is the
Mamdani method [23]. This method simply takes the
minimum strength of the rules in the inference system. For
example, in the rule IF Temperature is VERY COLD AND Door
State is OPEN THEN Run Time is VERY LONG. If the VERY
COLD and OPEN fuzzy values are 0.7 and 0.4, the VERY
LONG is assigned the minimum value which is 0.4. The
output of this stage is a strength value of each rule, which is
used in the next stage which involves defuzzification to
obtain the crisp output value [3, 24].
Fig -3: The structure of the ANFIS network.
The ANFIS structure consists of 5 different layers. The first
layer is the membership function layer, in which the input
variables are fuzzified via the membership function equation
to produce the membership degree value. The second layer
uses the output of the first layer (the membership degree
values) to compute the firing strength of each linguistic
variable in the second layer, which will be required for the
rules evaluation by the subsequent layers of the network
[29]. The third layer is the normalization layer, in which the
weights are normalized, this makes the value of each weight
in the range [0, 1]. The fourth layer carries out the rules
evaluation to determine the strength of each rule to the
overall output. The fifth layer uses a summation method for
the values obtained from the previous layer to determine the
final output value [30].

2.2.2. Model training and validation
To train and validate the ANFIS model, the overall data was
divided into two different sets, namely,thetrainingset which
was used for training and the model and the test set which
was used for evaluation of the model performance [31]. The
training process of the ANFIS involves a running of the
network for several epochs (cycles), during each epoch, the
error of the network prediction is computed via a cost
function and the weights of the network are updated to
minimize the error [32]. The root means squared error
(RMSE) was used as the cost function [33]. To determine the
optimum number of epochs, the networks were allowed to
run for 1000 epochs and the error was computed after each
epoch. The number of epochs that corresponds to the lowest
error is considered to be the optimum epoch number for
training the network.
After training the model, the test set was used to
assess the ability of the model to make predictions. The
model was applied to predict the run time of the test set, and
the error between the model’s prediction and the real value
was measured. The coefficient of determination (R2), the
MAE and the RMSE were calculated, which are commonly
used validation parameters to assess the performance of
machine learning models [22, 33, 34].
3.2 Hardware implementation
Considering that the purpose of the developed software is to
control air cooling devices, the developed ANFIS model was
installed and tested on a Field-Programmable Gate Array
(FPGA) embedded system [35]. The hardware part of the
system is composed of Zynq-7020 FPGA processor with the
AXI BRAM memory and AXI GPIO, these components are
available in the Vivado integrated development environment
(IDE), the hardware structure is shown in Figure 4.
Fig-4 The structure of the designed hardware.
The hardware architecture includes two ports, a memory
block RAM with 64 KB size and BRAM Controller. BRAM data
is accessed within single cycle latency. Advanced Xilinx
Interconnects (AXI) is introduced to the system hardware to
act as bus system. The general purpose input/output (GPIO)
unit provides a hardware interface between the processor
system AXI and the sensors (temperature and door state
sensors) [35, 36].
For implementing the software part of the system, the
software development kit (SDK) tools were used, which are
available in the Vivado IDE [37].
3.RESULTS AND DISCUSSION
3.1. Fuzzy logic system
The three built fuzzy logic systems were evaluated by
predicting the output variable (run time) using the
temperature and door state as input variables. The output of
each system was compared to the optimum run time and the
error of prediction was measured. The error plot of the three
systems is depicted in Figure 5. As can be seen, the triangular
membership function showed the lowest MAE value (4.766),
which indicates a lower prediction error compared to the
other two membership functions. The trapezoidal and
Gaussian membership functions had higher MAE values
(4.812 and 5.025, respectively) which reflects lower
prediction accuracy compared to the triangular membership
function.
Chart -1: membership function (MAE)
Considering that the triangular and trapezoidal
membership functions have similar shapes as well as input
parameters compared to the Gaussian membership function
[18, 19], it can be seen that they are more suitable for this
type of application, as they both showed lower and similar
error rate. The triangular membership function was further
used as the membership function in the ANFIS model
development as it displayed better prediction ability in
comparison to the other two membership functions.
3.2. ANFIS model
The ANFIS model was built using the triangular membership
function and the set of rules previously described. The
optimum number of epochs was determined to be 300
epochs based on the RMSE values. Figure 6 shows the change
in the RMSE values as the number of epochs progresses. As
can be seen, initially the RMSE value drops fairly quickly as
the model is learning the data, then at around epoch 200 the

The model obtained at 300 epochs was validated by
assessing its ability to predict the output variable (run time)
of the test set compounds. The validation parameters were
computed for the training and the test set compounds.
However, the training set validation parameters cannot be
considered reliable as the model was fit and trained using
this set. On the other hand, the validation parameters of the
test set are considered reliable as the data was not involved
in the training of the model, and reflects the model’s ability
to make predictions on unseen data. Table 3 shows the
validation parameters and their corresponding values. The
R2 value for both the training and test set was close to the
optimum value of 1, which shows that the model can explain
the variance in the output variable using the input variables.
The MAE values for the training and test set were 0.338 and
0.413, respectively, which are both low and indicate the
model can make accurate predictions. As expected the
training set MAE value was slightly lower because the model
was fit using the training set data and hence the lower error
is expected [34]. The RMSE values showed a similar trend to
the MAE values. Figure 7 shows the predicted output by the
model against the actual value for both the training and test
set.
Table-3: The validation parameters of the obtained ANFIS
model.
MEA(test)
0.413
RMSE
0.529
Parameter Value
R2 0.993
MAE 0.338
RMSE
0.424
R2(test)
0.992
RMSE value becomes more stable and drops more slowly. At
around epoch 300 the optimum number of epochs is reached
as the RMSE value begins to increase after this epoch, which
indicates the occurrence of over-fitting of the model due to
large number of epochs. The RMSE value continues to
increase afterward which is expected due to over-training of
the network.
Chart -2: The RMSE values were plotted against the epoch
number during the training of the ANFIS model.
Chart -3: The predicted output values by the ANFIS model
are plotted against the real values.
Comparing the ANFIS model performance with the fuzzy
logic system shows that the ANFIS model is significantly
superior in terms of accurate prediction. For instance, the
MAE values of the ANFIS model and the best obtained fuzzy
logic system (the triangular membership function system)
are 0.338 and 4.766, respectively, which reflects the higher
accuracy of the ANFIS model. This can be attributed to the
fact that the ANFIS model can learn the data and adjust to it
to make better predictions due to its neural networks
paradigm, which is more flexible comparedto the mere fuzzy
logic systems that use only a set of fixed rules evaluation to
estimate the output value [8, 10, 27].

[1] L. A. Zadeh, "Fuzzy logic," Computer, vol. 21, no. 4, pp.
83-93, 1988.
[2] F. M. McNeill and E. Thro, Fuzzy logic: a practical
approach. Academic Press, 2014.
[3] K. Tanaka and K. Tanaka, An introduction to fuzzy
logic for practical applications. Springer, 1997.
[4] J. C. Mugisha, B. Munyazikwiye, and H. R. Karimi,
"Design of temperature control system using
conventional PID and Intelligent Fuzzy Logic
controller," in 2015 International conference on fuzzy
theory and its applications (iFUZZY), 2015: IEEE, pp.
50-55.
[5] S. Basu, "Realization of fuzzy logic temperature
controller," International Journal of Emerging
Technology and Advanced Engineering, vol. 2, no. 6,
pp. 151-155, 2012.
[6] A. D. M. Africa, A. M. J. L. Divino, and K. A. B. Hartigan-
Go, "Fuzzy Logic Temperature Control: A feedback
control system implemented by fuzzy logic,"
International Journal, vol. 8, no. 5, 2020.
[7] J.-S. Jang, "ANFIS: adaptive-network-based fuzzy
inference system," IEEE transactions on systems,man,
and cybernetics, vol. 23, no. 3, pp. 665-685,1993.
[8] D. Karaboga and E. Kaya, "Adaptive network based
fuzzy inference system (ANFIS) training approaches:a
comprehensive survey," Artificial Intelligence
Review, vol. 52, no. 4, pp. 2263-2293, 2019.
[9] T. Mote and S. Lokhande, "Temperature control
system using ANFIS," International Journal of Soft
Computing and Engineering, vol. 2, no. 1, pp. 2231-
2307, 2012.
[10] S. Soyguder and H. Alli, "An expert system for the
humidity and temperature control in HVAC systems
using ANFIS and optimization with Fuzzy Modeling
Approach," Energy and Buildings, vol. 41, no. 8, pp.
814-822, 2009.
REFERENCES
4. CONCLUSION
The fuzzy logic system and ANFIS approaches were
conducted to develop fuzzy logic systems and an ANFIS
model to predict the optimum run time of air conditioners
using the current temperature and door state as input
variables. Three different membership functions were
examined and their performance was assessed. The
triangular membership function displayed better results
compared to the trapezoidal and Gaussian membership
functions. The obtained ANFIS model showed superiority
over the fuzzy logic systems in terms of prediction accuracy
as demonstrated by the lower error rates. Overall, the
ANFIS model proved to be highly efficient and accurate
for application in run-time control of air-conditioning
systems.
[11] R. Al-Jarrah and M. Al-Jarrah, "Developed adaptive
neuro-fuzzy algorithm to control air conditioning
system at different pressures," International Journal
of Engineering, Science and Technology, vol. 5, no. 4,
pp. 43-59, 2013.
[12] S. Van Der Walt, S. C. Colbert, and G. Varoquaux, "The
NumPy array: a structure for efficient numerical
computation," Computing in science & engineering,
vol. 13, no. 2, pp. 22-30, 2011.
[13] J. D. Hunter, "Matplotlib: A 2D graphics environment,"
Computing in science & engineering, vol. 9, no. 03, pp.
90-95, 2007.
[14] F. Herrera, E. Herrera-Viedma, and L. Martínez, "A
fuzzy linguistic methodology to deal with unbalanced
linguistic term sets," IEEE Transactions on fuzzy
Systems, vol. 16, no. 2, pp. 354-370, 2008.
[15] Z. Xu, "Group decision making with triangular fuzzy
linguistic variables," in International Conference on
Intelligent Data Engineering and Automated
Learning, 2007: Springer, pp. 17-26.
[16] M.-S. Chen and S.-W. Wang, "Fuzzy clustering
analysisfor optimizing fuzzy membership functions,"
Fuzzy sets and systems, vol. 103, no. 2, pp. 239-254,
1999.
[17] W. Pedrycz, "Why triangular membership
functions?," Fuzzy sets and Systems, vol. 64, no. 1,
pp. 21-30, 1994.
[18] O. A. M. Ali, A. Y. Ali, and B. S. Sumait, "Comparison
between the effects of different types of membership
functions on fuzzy logic controller performance,"
International Journal, vol. 76, pp. 76-83, 2015.
[19] A. Barua, L. S. Mudunuri, and O. Kosheleva, "Why
trapezoidal and triangular membership functions
work so well: Towards a theoretical explanation,"
2013.
[20] J. Zhao and B. K. Bose, "Evaluation of membership
functions for fuzzy logic controlled induction motor
drive," in IEEE 2002 28th Annual Conference of the
Industrial Electronics Society. IECON 02, 2002, vol. 1:
IEEE, pp. 229-234.
[21] A. M. Ajofoyinbo, V. O. Olunloyo, and O. Ibidapo-Obe,
"On development of fuzzy controller: The case of
gaussian and triangular membership functions,"

Journal of Signal and Information Processing, vol. 2,
no. 04, p. 257, 2011.
[22] C. J. Willmott and K. Matsuura, "Advantages of the
mean absolute error (MAE) over the root mean
square error (RMSE) in assessing average model
performance," Climate research, vol. 30, no. 1, pp. 79-
82, 2005.
[23] A. Fernández and F. Herrera, "Linguistic fuzzy rules in
data mining: follow-up Mamdani fuzzy modeling
principle," in Combining Experimentation and
Theory: Springer, 2012, pp. 103-122.
[24] O. Cordón, "A historical review of evolutionary
learning methods for Mamdani-type fuzzy rule-based
systems: Designing interpretable genetic fuzzy
systems," International journal of approximate
reasoning, vol. 52, no. 6, pp. 894-913, 2011.
[25] J. J. Saade and H. B. Diab, "Defuzzification techniques
for fuzzy controllers," IEEE Transactions on Systems,
Man, and Cybernetics, Part B (Cybernetics), vol. 30,
no. 1, pp. 223-229, 2000.
[26] F. Smith and Q. Shen, "Selecting inference and
defuzzification techniques for fuzzy logic control," in
UKACC International Conference on Control'98 (Conf.
Publ. No. 455), 1998, vol. 1: IET, pp. 54-59.
[27] M. Elsisi, M.-Q. Tran, K. Mahmoud, M. Lehtonen, and
M. M. Darwish, "Robust design of ANFIS-based blade
pitch controller for wind energy conversion systems
against wind speed fluctuations," IEEE Access, vol. 9,
pp. 37894-37904, 2021.
[28] O. Opeyemi and E. O. Justice, "Development of neuro-
fuzzy system for early prediction of heart attack,"
International Journal of Information Technology and
Computer Science, vol. 4, no. 9, pp. 22-28, 2012.
[29] O. C. Abikoye, E. O. Popoola, T. O. Aro, and V. O.
Popoola, "Adaptive Neuro-Fuzzy Inference System for
HIV/AIDS Diagnosis, Clinical Staging and Regimen
Prescription," Computer Science &
Telecommunications, vol. 51, no. 1, 2017.
[30] A. Abdel-Aleem, M. A. El-Sharief, M. A. Hassan, and M.
G. El-Sebaie, "Implementation of fuzzy and adaptive
neuro-fuzzy inference systems in optimization of
production inventory problem," Applied Mathematics
& Information Sciences, vol. 11, no. 1, pp. 289-298,
2017.
[31] J.-S. Jang, "Input selection for ANFIS learning," in
Proceedings of IEEE 5th international fuzzy systems,
1996, vol. 2: IEEE, pp. 1493-1499.
[33] W. Wang and Y. Lu, "Analysis of the mean absolute
error (MAE) and the root mean square error (RMSE)
in assessing rounding model," in IOP conference
series: materials science and engineering, 2018, vol.
324, no. 1: IOP Publishing, p. 012049.
[35] M. R. Khalil and L. A. Mohammeed, "Embedded
processor system for controllable period-width
multichannel pulse width modulation signals,"
TELKOMNIKA (Telecommunication Computing
Electronics and Control), vol. 19, no. 1, pp. 220-228,
2021.
[36] G. Nhivekar, S. Nirmale, and R. Mudholker,
"Implementation of fuzzy logic control algorithm in
embedded microcomputers for dedicated
application," International Journal of Engineering,
Science and Technology, vol. 3, no. 4, 2011.
[37] T. Feist, "Vivado design suite," White Paper, vol. 5, p.
30, 2012.
[32] L. Naderloo et al., "Application of ANFIS to predict
crop yield based on different energy inputs,"
Measurement, vol. 45, no. 6, pp. 1406-1413, 2012.
[34] H. Dette and A. Munk, "Validation of linear regression
models," Annals of Statistics, pp. 778-800, 1998.

Estimation of Air-Cooling Devices Run Time Via Fuzzy Logic and Adaptive Neuro-Fuzzy Inference System

More Related Content

Similar to Estimation of Air-Cooling Devices Run Time Via Fuzzy Logic and Adaptive Neuro-Fuzzy Inference System (20)

More from IRJET Journal (20)

Recently uploaded (20)

Estimation of Air-Cooling Devices Run Time Via Fuzzy Logic and Adaptive Neuro-Fuzzy Inference System