 Open Access
 Total Downloads : 212
 Authors : Siham A. M. Almasani, Wadeea A. A. Qaid , Ahmed Khalid, Ibrahim A. A. Alqubati
 Paper ID : IJERTV4IS080702
 Volume & Issue : Volume 04, Issue 08 (August 2015)
 Published (First Online): 07092015
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Fuzzy Expert Systems to Control the Heating, Ventilating and Air Conditioning (HVAC) Systems
Siham A. M. Almasani1 Wadeea A. A. Qaid1 Ahmed Khalid2 Ibrahim A. A. Alqubati2
1. Hodeidah University, Faculty of computer science and engineering, Yemen
2. Najran University, Community College ,Computer Department, KSA
Abstract – This paper presents a fuzzy expert system for heating, ventilating and airConditioning (HVAC) system to provide a comfortable environment in terms of humidity, temperature and other environmental parameters for the occupants as well as to save energy. The current values of temperature, humidity and oxygen inside the building and the temperature outside the building is used as inputs variables for the introduced system, and the Heat valve, Cold valve, Speed of exhaust motor and Speed of water pump as output variables. Fuzzy logic Toolbox and Simulink in LABVIEW software are used to simulate the proposed system. The system shows good performance in controlling the HVAC system.
Keywords: Fuzzy Expert System, control, HVAC, Nonlinear system, PID, Autotuning PID.

INTRODUCTION.
Commercial and industrial HVAC systems use electric and mechanical control system to maintain the desired temperature humidity level, and static pressure within a given area or zone [10]. Previous work have established a foundation of process models and control strategies suitable for building HVAC systems. Good results are obtained with application of PID autotuning controllers [20], neural network techniques [25], and even genetic algorithms [17]. Although a great deal of research has been conducted to improve the performance of HVAC control system [21, 7, 15]. In [19] Hybrid PIDCascade Control for HVAC System was proposed. In recent years, researchers have extensively used the fuzzy logic for modeling, identification, and control of highly nonlinear dynamic systems [16,27]. To decrease the energy consumption researchers in the area of thermal comfort [12,6,18] have learned that the required indoor temperature
of a building is not a fixed value. In fact, the certain range of temperatures is sufficient to create a comfortable situation. Therefore, the ideal HVAC system works with high efficiency supplying only the amount of heat, cold and air that is necessary to maintain internal conditions at a level providing thermal comfort to room occupants.
Fuzzy logic like human logic has no limits, which based on decisionmaking methods. Therefore, to make a better decision, controlling the operation is needed which has in turn led to use fuzzy control mechanism that is based on logic [8]. Fuzzy logic controller systems do not require full knowledge of the model, while in another known controller this knowledge is required [9]. The uncertainty tests of fuzzy systems and experts knowledge are very important as control has become popular and used in many different fields of science [22, 24, 5]. Fuzzy logic gave the best result when it used to control heating system [11].
In this article, we use fuzzy expert system. Fuzzy logic is a branch of artificial intelligence that deals with reasoning algorithms used to emulate human thinking and decision making in machines. FLCs are suitable for control of HVAC system because their inputs and outputs are real valued variables, mapped with a nonlinear function. These kinds of systems achieve an alternative for those applications where classical control strategies do not achieve good results. In many cases, these systems have two characteristics: the need for human operator experience, and a strong non linearity, where it is not possible to obtain a mathematical model [23,14,2].
The basic configuration of the fuzzy logic system consists of four main components: fuzzifier, fuzzy inference engine, fuzzy rule base, and defuzzifier.
Knowledge Base
Fuzzification Interface
Inference System
Defuzzification Interface
Controlled System
State Variables Control Variables
Fig. 1. Generic structure of fuzzy logic controller.

HVAC FUZZY EXPERT SYSTEM
Fuzzy logic is one of the methods of Soft Computing. Soft Computing is a computational method that is tolerant to suboptimality, impreciseness, vagueness and thus giving quick, simple and sufficient good solutions [4]. Fuzzy logic together with the appropriate rules of inference provides a power framework for managing uncertainties pervaded in medical diagnosis and control [13]. Fuzzy logic system take the input which may be imprecise and vague and
applied reasoning that is approx., rather than precise and as a result give their decision [1]. There are two concepts within fuzzy expert system that play an important role in its application. The first is the linguistic variable, that is, a variable whose values are words or sentences in a natural synthetic language. The second is a fuzzy ifthen rule in which the antecedent and consequent are propositions containing linguistic variables, which exploits the tolerance for imprecision and uncertainty [26]. Fig. 2 represent the HVAC expert system Block Diagram system
Input Temperature, humidity, oxygen and outdoor Temperature
Fuzzy Expert System
Fuzzy Rule Base,
Memberships Function
heat valve, cold valve, speed of exhaust motor and speed of water motor
Room temperature, humidity, oxygen
Fig. 2 the Block Diagram of HVAC System
Table 1. Shows the input variables and their linguistic values for the HVAC fuzzy expert system.
Table. 1. The input variables and their linguistic values.
Parameters 
Linguistic values 
Temperature 
Very Low, Low Medium, High, Very High 
Humidity 
Very Low, Low, Medium, High 
Oxygen 
Very Low, Low, Medium, High 
Outdoor Temp 
Cold ,Hot 
Table 2. Shows the output variables and their linguistic values for the HVAC fuzzy expert system.
Table.2. The output linguistic variables and their values
Parameters 
Linguistic values 
Heat valve 
Off, Low, Medium, High 
Cold valve 
Off, Low, Medium, High 
Speed of exhaust motor 
Low, Medium, High 
Speed of water pump 
Off, Low, Medium, High 
In this paper Mamdani's fuzzy inference method is used to build fuzzy IFTHEN rules because it simple structure for 'minmax' operations [2, 3].
The fuzziness methods for Temperature, Humidity and oxygen is Gaussian and the membership function is shown in equation (1) (Temperature=X, temperature value=x)
f X : , c e
x c2
2 2
(1)
For heat valve motor, cold valve motor, speed of exhaust motor and speed of water pump the membership function in triangle fuzziness methods as shown in equation (2) (Heat valve =Y, value of speed heat valve =y).
0,
y a
c y
f Y :a, b, c b a
c b
0
y a
a y b
b y c c y
(2)
The parameters a is a locating of the feet and c is a locating of the peak. In the equation (2).
The fuzzy set for the linguistic expressions and membership functions of both input and output are obtain as seen in Fig 67.
Fig. 3 The fzzy set for the linguistic expressions and membership functions of input parameters
Fig. 4. The fuzzy set of the linguistic expressions and membership functions of the output parameters
The Center of Area Defuzzification method is used in the proposed system to convert the degrees of membership of output linguistic variables into numerical values which is compound with Mamdani's fuzzy inference method as shown in equation (3)
xmax
f x xdx
x
CoA xmin
max
(3)
xmin
f xdx
Table 3. Shows some of the rules composed by the expert for heating, ventilation and air condition the total of rules for the proposed fuzzy expert system is 100 rules.
Rules
Temperature
Humidity
Oxygen
Outdoor Temp
Heat valve motor
Cold valve Motor
Water pump speed
Exhaust motor speed
Table. 3. Some of the rules for heating, ventilation and air condition

Low Low 
. Low 
Low Low 
Cold Then Hot Then 
Off High 
Low Off 
. Low 
High High 
3.if 
VLow 
.. 
Low 
Cold Then 
Off 
Med 
. 
Med 
4.if 
VLow 
Low 
Low 
Hot Then 
Low 
Off 
High 
High 
5.if 
Med 
.. 
Low 
Cold Then 
Off 
High 
Low 

.. 
.. 
.. 
.. 
.. 
.. 
.. 
.. 
.. 
90.if 
VHigh 
Med 
Low 
Hot Then 
Med 
Off 
Low 
High 
91.if 
VHigh 
High 
Low 
Hot Then 
Med 
Off 
Med 
High 
92.if 
VHigh 
Vlow 
VLow 
Hot Then 
Med 
Off 
Low 
Low 
93.if 
VHigh 
Med 
VLow 
Hot Then 
Med 
Off 
Low 
Low 
94.if 
VHigh 
High 
VLow 
Hot Then 
Med 
Off 
Med 
Low 
95.if 
VHigh 
Vlow 
Med 
Hot Then 
Med 
Off 
Low 
Low 
.. 
.. 
.. 
.. 
.. 
.. 
.. 
.. 
.. 
3.2. HVAC expert system Mathematical Model
The mathematical model is very important to understand how the system behaves. Equation 4 through 16 represent the mathematical model for the heating, air conditioning, humidity and ventilation systems.
The heat system mathematical model is determined by relating the heat flow into the room from the heater to the room temperature. The amount of loss of heat air from the heater and the room temperature using the following equations:
dtroom
1 dQFlow Heat dQlosses
(4)
dt M c dt dt
dQ Troom Toutdoor
(5)
dt R
Losses eq
Where: M mass of air inside the house; Req equivalent thermal resistance of the house; Heater; dQlosses / dt heat losses to the environment.
Now determine the heat flow from the Heater expressed by the equation (6):
dQ
dQFlow Heat / dt heat flow from the
Theater pv M f c
dt
(6)
Where:
dQ / dt heat flow from heater in the room; Theater largest temperature coming from the heater; pv the power of
heat valve; Mf air mass flow rate through heater; c heat capacity of air at constant pressure.
The Air Conditioning mathematical model is determined by relating the cool flow into the room from air condition to the room temperature, the amount of loss of cold air from air condition and the room temperature using the following equations:
dtroom 1 dQFlowCool

dQlosses
(7)
dt M c
dt dt
( dQ )
dt
losses
T T
outdoor room
Req
(8),
Where: M mass of air inside the house; Req equivalent thermal resistance of the house; conditions; (dQ / dt)losses cool losses to the environment.
Now determine the cool flow from the Air condition expressed by the equation (9):
dQ
dQFlowCool / dt cool flow from air
TCool pv M f c
dt
(9),
Where: dQ / dt cool flow from air condition in the room; Tcool largest cool degree coming from the air conditioner; pv the power of cold valve; Mf air mass flow rate through air conditioner; c heat capacity of air at constant pressure.
The Humidity mathematical model is determined by relating the flow of steam into the room from humidity to the
room humidity, the amount of loss of steam air from humidity and the room humidity using the following equations:
dhroom
bs
dQFlowSteam dQlosses
(10)
dt 1000
dt dt
( dQ ) H H R dt losses outdoor room
(11),
where: bs the size of the house (height*width*length);
dQFlowSteam / dt the amount of steam flow from the humidity for
1000 cubic feet of the size of the place to be moistened; R equivalent steam leaks of the house; (dQ / dt)losses steam losses to the environment.
Now determine the steam flow from the humidity is express by the equation (12):
dQ
Hsteam pv M
dt
(12),
where:
dQ / dt steam flow from humidity in the room; largest steam degree coming from the moisturizer; pv
power of steam valve; M steam mass flow rate through moisturizer;
The Ventilation mathematical model is determined by relating the flow of oxygen into the room from ventilation to the room oxygen and the amount of loss of oxygen air from ventilation as following:
Sp ( dQFlowOxygen
1 dQ
) losses fo
(13)
dt dA dt
dvroom
Sp
100
(14)
dt
( dQ )
dt
losses
vol ACH
FR /1000
(15),
where: Sp the amount of oxygen flow from the ventilation after subtracting the missing oxygen; dA density of the Air; dQFlowOxygen / dt the amount of oxygen flow from ventilation; (dQ / dt)losses oxygen losses to the environment; fo fraction of oxygen in air, it is equal 0.2095; vol size of the house (height*width*length); ACHthe air change rate; equivalent oxygen leaks of the house.
Now determine the oxygen flow from ventilation expressed by the equation (16):
dQ
dt
Voxygen pv
(16),
Where: dQ / dt oxygen flow from ventilation in the room; largest oxygen degree coming from the ventilation; pv
power of speed of ventilation.
4. RESULTS AND DISCUSSION
Fuzzy logic Toolbox and Simulink in LABVIEW software used to design a simulation system of HVAC fuzzy expert system controller for heating, ventilation andAir Condition control System. Fig.5 shows the interface for the proposed system.
Fig.5 HVAC fuzzy expert system Interface
Fig. 6 shows the results when the temperature = 70Â°F, humidity=50%, oxygen=20%, outdoor humidity=20 and outdoor temperature = 45Â°F
Fig. 6. A result of using Fuzzy Expert Controller In the condition of working heating, humidifier and ventilation
From Fig.6, we see the outdoor temperature is lower than the desired temperature; the system will open the heat valve to heat the place and close the cold valve. As long as the required humidity is greater than the outdoor humidity the humidifier will be open and the exhaust motor is operating, so the system is working to give us the required comfort without overshoot.
Fig. 7 shows the results when the temperature = 70Â°F; set humidity=60%, set oxygen=25%, outdoor humidity=30 and outdoor temperature = 100Â°F.
Fig. 7. A result of using Fuzzy Expert Controller in the condition of working air condition, humidifier and ventilation
From Fig.7 we see the outdoor temperature is higher than the desired temperature, the system will open the cold valve to cool the place and close the heat valve, also the required humidity is greater than the outdoor humidity thus the humidifier and exhaust motor is working, so we note that the system is working and also gives us the required comfort without overshoot.
From above results, the simulation provided good performance in terms of oscillations and overshoot in the absence of prediction mechanism.
The result of using the fuzzy expert system has reduced overshoot and steady state error and significant improvement in maintaining performance over the widely, compared with traditional control like PID control, auto tuning PID control and on/off control the fuzzy expert system is the best.
In the result of examining the robustness of these controllers with respect to external disturbances, the fuzzy controller shows better control performance than PID controller in terms of settling time, overshoot and rise time as shown in table 4
Table. 4. Performance characteristics of HVAC system with Fuzzy and PID controllers
Controller 
Overshoot 
Rise time 
Settling Time 
PID 
8.5 
0.009 
43.33 
Fuzzy 
2.25 
0.001 
2.28 
6. CONCLUSION
In this paper, the real time implementation results show that the approach of using the fuzzy logic control with the expert system is the best way to control system and gives good results, the maximum overshoot of the system is 2.25% but in the conventional controller like PID control the maximum overshoot of the system is higher. Therefore, the performance of fuzzy expert controller is better than of conventional algorithm.
We found that the heating, ventilation and air condition system with the Fuzzy expert controller provides the accurate control. The fuzzy controller response in all experimental conditions is quite as expected, and it does not exceed the maximum and minimum for required limit. The FLC algorithm adapts quickly to longer time delays, provides a stable response and can give more attention to various parameters, such as time of response, error of steadying and overshoot.
REFERENCES

Anjali Baghel1, Tilotma Sharma2, Survey on Fuzzy Expert System, International Journal of Emerging Technology and Advanced Engineering, Volume 3, Issue 12, December 2013.

Bolata, F., Nowr, A. From fuzzy linguistic specifications to fuzzy controllers using evolution strategies, IEEE International Conf. on Fuzzy Systems, pp.10891094. 1995

Bova, S., Codara, P., Maccari, D., Marra, V. A., logical analysis of Mamdanitype fuzzy inference, I theoretical bases, IEEE International Conference on Fuzzy Systems,
Barcelona, pp. 18, 2010

Chen, C.L. and Chen, W.C., Fuzzy Controller Design by Using Neural Network Techniques. IEEE Transactions on Fuzzy Systems. 2(3):235244, 1994.

Hagras, H., Callaghan, V., Colley, M., Clarke, G., A., hierarchical fuzzygenetic multiagent architecture for intelligent buildings online learning, adaptation and control. Elsevier, Information Sciences, pp. 3357, 2003.

Hamdi, M., Lachiver, G., A fuzzy control system based on the human sensation of thermal comfort. IEEE, International Conf. on Fuzzy System, Washington DC., USA, pp. 49, 1998.

Hartman, Thomas, New horizons for HVAC control, Heating Piping Air Conditioning, Vol. 67, No. 3, 1995, pp. 3341.

Herrod, R.A.; Papas, B.C., Industrial applications of artificial intelligence. IEEE, Pulp and Paper Industry Technical Conference, pp. 8690., 1989.

Jian, W., Wenjian, C., Development of an adaptive neuro fuzzy method for supply air pressure control in HVAC system. IEEE, International Conf. on Systems, Man, and Cybernetics, vol. 5, pp.38063809., 2000.

Jianbo Bai, "Development an Adaptive Incremental Fuzzy PI Controller for a HVAC System", INT J Computer, Communication and Control, IDDN 18419836 8(5), pp. 654661, 2013

Kobersi, I.S., Finaev, V.I., Almasani, S.A., Kaid, W.A.A., Control of the Heating System with Fuzzy Logic, World Applied Sciences Journal 23 (11): 14411447, 2013ISSN 18184952, 2013.

Liang, J., Du, R., Design of intelligent comfort control system with human learning and minimum power control strategies, Elsevier Science, Energy Conversion and Management, vol. 49, pp. 517528, 2008.

Ludmila, I.K. and Steimann F., Fuzzy Medical Diagnosis. School of Mathematics, University of Wales: Banggor, UK. 2008.

Mamdani, E.H., Applications of fuzzy algorithms for control a simple dynamic plant, in Proc. of the IEEE, vol. 121, no. 12,pp. 15851588, 1974.

Mulcahy, Robert W, New developments in HVAC control systems and the impact ofdirect digital control technology, Australian refrigeration, Air conditioning and heating, Vol. 38, No. 12, 1984, pp. 2419.

Nguyen, H. T., Prasad, N. R., Walker, C. L., and Walker, E. A., A First Course in Fuzzy and Neural Control, USA: Chapman & Hall/ CRC, 2003.

N. NASSIF, S. KAJL and R. SABOURIN: Evolutionary algorithms for multiobjective optimization in HVAC system control strategy. In IEEE Annual Meeting of the North American Fuzzy Information Processing Society – NAFIPS, 2004., 1 (2004), 5156

Nowak, M., Urbaniak, A., Utilization of intelligent control algorithms for thermal comfort optimization and energy saving. IEEE, Carpathian Control Conf. (ICCC), pp. 270 274, 2011.

Raad Z. Homod, Khairul Salleh Mohamed Sahari, Haider A.
F. Mohamed, Farrukh Nagi, "Hybrid PIDCascade Control for HVAC System", International Journal of Systems Control (Vol.12010/Iss.4), pp. 170175.

Q.G. WANG, C.C. HANG Y. ZHANG and Q. BI: Multivariable controller autotuning with its application in HVAC systems. In Proc. of the American Control Conference, 1999, San Diego, California, USA, 6 (1999), 43534357.

Soyguder Servet, Karakose Mehmet, Alli Hasan, Design and simulation of selftuning PIDtype fuzzy ada`ptive control for an expert HVAC system, Expert Systems with Applications, Vol. 36, No. 3, 2009, pp. 45664573.

Wakami, N., Araki, S., Nomura, H., Recent applications of fuzzy logic to home appliances. In Proceedings of the IECON93. International conference on industrial electronics, Vol. 1, pp. 155160., 1993.

Wang, H.O., Tanaka, K., Griffin, M.F., An approach to fuzzy control of nonlinear systems: stability and design issues,
IEEE, Fuzzy Systems, pp. 1423, 1996

Wong, J.K.W., Li, H., Wang, S.W., Intelligent building research: a review. Elsevier, Automation in Construction. 14:pp. 143 159, 2005.

Y. CHN and Z. CHEN: A neuralnetworkbased experimental technique for determining Ztransfer function coefficients of a building envelope. Building and Environment, 35(3), 2000, 181189.

ZADEH, L.A.: Outline of the new approach to the analysis of complex systems and decision processes, IEEE Trans. Syst., Man Cybern., Jan. 1973, 3, (l), pp. 2844.

Z. Kovaci c, Bogdan, S., Crnoija, P., An automated fuzzy Â´ controller design for highorder systems, AutomatikaLtd., vol. 2, no. 3, pp. 3740, 1996.