Microsoft Word - 17-2871_s_ETASR_V9_N4_pp4405-4410


Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4405  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

Greenhouse Modeling, Validation and Climate 

Control based on Fuzzy Logic 
 

Manel Jomaa 

UR-LAPER, Department of Physics, 
Faculty of Sciences of Tunis, University of Tunis El Manar, 

Tunis, Tunisia 

manel.jomaa@enit.utm.tn 

Fernando Tadeo 

Departamento de Ingenieria de Sistemas y Automatica,  

Institute of Sustainable Processes, University of Valladolid, 
Valladolid, Spain 

fernando.tadeo@uva.es 

Mehdi Abbes 

Laboratoire Analyse et Commande des Systemes, 
National Engineering School of Tunis, 

Tunis, Tunisia 

mehdi.abbes@enit.rnu.tn 

Abdelkader Mami 

UR-LAPER, Department of Physics, 

Faculty of Sciences of Tunis, University of Tunis El Manar, 
Tunis, Tunisia 

abdelkader.mami@fst.utm.tn  
 

 

Abstract—This paper deals with the modeling and control of the 

air temperature and humidity in greenhouses. A physical model 

of the greenhouse used in the Simulink/Matlab environment is 

elaborated to simulate both temperature and indoor humidity. As 
a solution to the non-linearity and complexity of the greenhouse 

system, a fuzzy logic method is developed to control the actuators 

that are installed inside the greenhouse for heating, ventilation, 
humidification and cooling to obtain a suitable microclimate.  

Keywords-greenhouse; fuzzy logic controller; air temperature; 

humidity; Simulink; Matlab 

I. INTRODUCTION  

Agricultural production is a central economy field in many 
countries. Tunisia is considered one of the best worldwide 

agricultural product exporters. Nowadays, greenhouse is the 

main solution to climate change effects. Its main purpose is to 
effectively control the environment and protect the plants by 
creating a favorable climate. Among the conditions in a 

greenhouse, this paper focuses on indoor temperature and 

humidity. An optimal vegetation yield can be produced from 

the precise configuration of these conditions [1]. Numerous 
methods of greenhouse control have been studied and 

developed. The regulations inside the greenhouse consist of 

evaporation, fogging and ventilation systems. This paper 
presents a control system useful for adaptation despite the non-

linearity and complexity of greenhouse dynamics [2, 3]. 

Various types of climate control greenhouse systems have been 
previously mentioned. In [4, 5], a feedback controller approach 
has been discussed. The type of climate control mentioned in 

[6] has been developed to the optimal control method. Neural 

networks were studied in [7-9]. Authors in [10, 11] offer a 
regulation of temperature and humidity inside the greenhouse 

with a fuzzy logic regulator. In [12], the Takagi-Sugeno 

method of the air temperature control is described. In [13-15], 

model predictive control (MPC) was used. The proposed work 

starts with the simulation of the dynamic greenhouse model 
followed by its validation and its control via Fuzzy Logic. 

II. GREENHOUSE MODEL 

The model of the proposed input-output system can be seen 

in [16]. The considered variables are: intT  is the interior 

temperature and intH  is the humidity variable. The adjustment 

of these variables will be obtained by αMV  which is the 
ventilation system, wMV  which is the heating system, and

fogMV  which is the fogging system. The measurable 

disturbances consist of wind speed, solar radiation, and outside 

humidity and temperature [16]. 

A. Proposed Model 

The first step of the proposed study is the development of 
the dynamic model of the greenhouse using the 
Simulink/Matlab. Many physical models like the energy 
balance and mass balance have been suggested to describe the 
dynamics of the greenhouse climate [17, 18]. 

B. Greenhouse Model with Simulink 

A derived simple model is required to design a control 
system that analyzes and guesses the correct variables of 
cooling, heating and ventilation systems. The model consists of 
conventional dynamic equations that correspond to the indoor 

air temperature iT , soil temperature mT , and indoor humidity 

iX [17]. The differential equations are given by the heat and 

humidity balance. 

1) Heat Balance 

The dynamic model of the air temperature and the thermal 
mass temperature inside the greenhouse are described by (1) 

Corresponding author: Manel Jomaa



Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4406  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

and (9) [17]: 

( )
dTi

v c Q Q Q C Q Q Q Wi p s cc m sat e n v
dt

ρ = − + − + − +
 

(1) 

where: 

sQ  is the solar energy supplied to the air volume 

determined by (2): 

0SiAsQ τ=      (2) 

ccQ  is the energy exchange by the conduction and the 

convection phenomena that are explained using (3): 

( )oTiTcAiAccQ −=     (3) 

mQ  is the energy exchange between the thermal mass and 

the inside air determined by (4): 

( )iTmTmhiAmQ −=      (4) 

eQ  is the energy loss due to the crop evapotranspiration: 

eQ Eλ=      (5) 

nQ  is the energy loss by the nebulization: 

nQ fogλ=      (6) 

vQ  is the energy exchange due to the window ventilation: 

( )v p i oQ c G T Tρ= −     (7) 

W  is the energy provided from the heating system: 

max
100

wMV
W W=     (8) 

The thermal mass temperature inside the greenhouse is 
described by (9): 

m
i m sm m f

dT
A C Q Q Q

dt
= − −    (9) 

Knowing that: 

sm m sQ Qα=      (10) 

Equation (10) is the energy stored by the thermal mass 
during the day and (11) is the energy loss through the ground: 

m ref

f i a
ref

T T
Q A k

Z

 −
=  

 
 

    (11) 

2) Humidity Balance  

The model of the relative humidity inside the greenhouse is 

determined by (12) [17]: 

( )ii v sat
dX

V F C E fog
dt

ρ = + +    (12) 

where: 

vF  is the water rate in the air renewal flow: 

( )v o iF G X Xρ= −     (13) 

E  is the crop evapotranspiration identified by (14): 

( )2

1

i n p iA R L c D gwb
E

gwb

gws

ρ

γ λ

∆ +
=

   
∆ + +       

   (14) 

fog  is the water rate of the fogging system: 

max
100

fogMV
fog fog

 
=  
 

   (15) 

The greenhouse that is used for the validation is situated at 

the C.R.T.EN Borj Cedria Park of science and technology in 

Tunisia, in a semi-urban environment. The greenhouse axes 

are in the East-West direction. The dimensions are 12.5m 

length, 8m width and 3m height. The acquisition of both 

external weather conditions (wind speed, temperature, 

radiation, and humidity) and internal climate greenhouse data 

are connected to a Campbell 21X data logger. The 

measurements of external and internal relative humidity are 

performed via HMP35C sensors with 3% precision and a 

range between 0% and 100%. The external and internal 

temperatures are measured in a range from -35°C to +50°C. 

The solar radiation is measured by a pyranometer with 5% 

precision [3]. Figure 1 shows the solar radiation measured in 

the 29th, 30th and 31st of March, 2001 whereas Figure 2 

presents the wind speed, measured at the same dates. 

 

 

Fig. 1.  Reached cover solar radiations, March, 29–31, 2001 

 
Fig. 2.  Wind speed March, 29–31, 2001 

Figures 3-5 present the simulation results, for the same time 
period, of internal air temperature (K), internal soil temperature 

(K) and relative humidity (%).  



Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4407  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

 
Fig. 3.  Simulation results of the air temperature inside the greenhouse 

 
Fig. 4.  Simulation results of the soil temperature inside the greenhouse 

 
Fig. 5.  Simulation results of the relative humidity inside the greenhouse  

The simulated results obtained via Simulink are compared 

with the detailed model discussed and described in [3]. The 
same parameters of the greenhouse are used in both cases. 

 

 
Fig. 6.  Measured and simulated internal air temperatures 

 
Fig. 7.  Measured and simulated results of internal air humidity 

Figure 6 presents the simulated and measured indoor air 
temperature. Figure 7 shows the simulated and measured 

results of the indoor relative humidity. It can be seen that a 

good agreement is obtained between the simulated and the 

measured results of both temperature and relative humidity. In 

order to activate the actuators, power should be supplied. In 
this case, a logic control is selected and its pseudo-code is 

briefly described in Figure 8. Figures 9-11 present the heating, 

ventilation and humidifying rates that are delivered by the 

control algorithm in Figure 8. 

 

 

Fig. 8.  Algorithm block diagram 

 

 
Fig. 9.  Heating rate 

 
Fig. 10.  Ventilation rate 

 
Fig. 11.  Humidifying rate 



Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4408  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

III. FUZZY LOGIC CONTROLLER 

Fuzzy logic has been considered an attractive solution for 

various engineering problems due to features such as its 

robustness, simplicity and flexibility. Three stages must be 

satisfied to design this type of control system: Fuzzification, 
inference rules and defuzzification [20]. The main aim of this 

system is to control the climate inside the greenhouse and boost 

its effectiveness.  

A. The Structure of The Control System 

The input variables of temperature and humidity in the 

fuzzy logic controller are represented by T∆ with outside 
temperature and H∆ with outside humidity knowing that:  

( ), , , ,T T T NB NM Z PM PBdesired indoor∆ = − ∈  (16) 

where outside temperature ∈(PM, Z, PG). 

( ), , , ,H H H NB NM Z PM PBdesired indoor∆ = − ∈  (17) 

and outside humidity ∈(PM, Z, GP) 

where    














BigPositivePB

MediumPositivePM

ZeroZ

MediumNegativeNM

BigNegativeNB

_:

_:

:

_:

_:

   









BigPositivePB

ZeroZ

MediumPositivePM

_:

:

_:

 

The methodology described in Figure 8 was used to 

develop this command. During the control we took into 
account the external environmental conditions (outside 
temperature and outside humidity). The output variables are the 

heating rate (Heating), the ventilation rate (Ventilation) and the 

humidifying rate (Humidifying). 

(Heating, Ventilation) ∈ (Zero, Medium, High) 

(Humidifying) ∈  (Zero, Medium, High) 

The fuzzy inference method that is used in this work is the 

Mamdani method and the defuzzification technique is based on 
the center of gravity method. 

B. Decoupling Fuzzy Logic Controller 

Figure 12 presents the input variables of the temperature 
and outside temperature and Figure 13 presents the input 

variables of the humidity and outside humidity. Figure 15 

presents the output variables of heating, ventilation and 

humidifying rates. 

IV. SIMULATION RESULTS 

The main objective of control is to smoothly regulate air 
temperature and humidity inside the greenhouse to their desired 

values ( desiredT , desiredH ). We controlled the actuators by 

lowering their rates in order to reduce energy. The value of the 
desired temperature is 15°C at night and 28°C at day. The 

desired relative humidity is 70%. The tracking performance of 

temperature and humidity inside the greenhouse are illustrated 
in Figures 15 and 16 during simulation time of one day.  

 

Fig. 12.  Membership functions of temperature error (∆T) and outside 

temperature 

 

Fig. 13.  Membership functions of humidity error (∆H) and outside 

humidity 

 

 

 

Fig. 14.  Membership functions of heating, ventilation and humidifying 

rates 



Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4409  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

 
Fig. 15.  Simulation results of the greenhouse temperature with fuzzy logic 

controller 

 
Fig. 16.  Simulation results of the greenhouse humidity with fuzzy logic 

controller 

The provided heating and ventilation rates are displayed in 

Figures 17 and Figure 18. A regulator fuzzy logic controller is 

applied to control the humidifying system in order to regulate 
the relative humidity at its set point value, as presented in 

Figure 19.  

 

Fig. 17.  Heating rate 

 
Fig. 18.  Ventilation rate 

 
Fig. 19.  Humidifying rate 

V. CONCLUSION 

In this article, a greenhouse model has been proposed and 

validated. A comparison between the simulated and measured 

results of the internal air temperature and the relative humidity 

has been made using the parameters proposed in [3]. A good 
agreement has been obtained between the calculated and 

measured data, which indicates the efficiency of the proposed 

model. Based on this model, a fuzzy logic controller has been 

developed to satisfy the demands in terms of temperature and 
humidity, using the measured data. The results showed a stable 
behavior of both temperature and humidity with a low rate of 

heating, ventilation and humidification. This behavior has been 
achieved without using the dehumidifier system in order to 

reduce energy consumption. In the future, we will focus on 

developing an improved control model for the greenhouse 

climate and testing the stability and robustness of the fuzzy 
logic controller with more variables representing the complete 

state of the greenhouse. In addition, we will design a power 

system based on renewable energy sources in order to reduce 

the use of grid utilities. 

NOMENCLATURE 

A [m²] Windows area [50m²] 

a Constant for renewal volumetric flow 

Ac Loss coefficient of conduction and convection 

Ai [m²] Greenhouse surface area [100m²] 

Cm Thermal mass heat capacity 

Cp [J Kg
-1  
°C

-1
] Air heat capacity [1003J·Kg

-1 
°C

-1
] 

Csat Air saturation coefficient 

Di [kpa] Air water vapor deficit 

E [kgH2O S
-1
] Crop evapotranspiration  

fog[kgH2O S
-1
] Water rate of fogging system 

Fv [kgH2O S
-1
] Water rate in the air renewal flow 

G [m
3
 S

-1
] Renewal air flow 

gwb [m
 
S
-1
] Boundary-layer conductance 

gws [m
 
S
-1
] Stomatal conductance 

hm [W m
-1
 K

-1
] Conductivity coefficient between air and thermal mass 

Hi [%] Inside relative humidity 

Ho [%] Outside relative humidity 

L Leaves area index 

MVfog [%] Fogging system manipulated variable  

MVw [%] Heating system manipulated variable  

MVα [%] Windows opening manipulated variable 

psat [kPa] Saturation pressure  

p [kPa] Atmospheric pressure [98.1kPa] 

Qcc [W] 
Energy exchange by conduction and convection 

phenomena 

Qe [W] Energy loss due to crop evapotranspiration 

Qf [W] Energy loss through ground 

Qm [W] Energy exchange with thermal mass 

Qn [W] Energy loss by nebulization 

Qs [W] Solar energy supplied to air volume 

Qsm [W] Energy stored by thermal mass during the day 

Qv [W] Energy exchange due to window ventilation 

Rn [W m
-2
] Solar radiation absorbed by the crop 

So [W m
-2
] Solar radiation 

Ti [°C] Inside temperature 

Tm [°C] Thermal mass temperature 

To [°C] Outside temperature 

Tref [°C] Ground temperature at reference depth 

Vi [m
3
] Greenhouse volume [300m

3
] 

V [ms
-1
] Wind speed 

W [W] Energy from the heating system 

Wmax [W] Maximum power of the heating system 

Xi [kgH2O kgair
-1
] Inside absolute humidity 



Engineering, Technology & Applied Science Research Vol. 9, No. 4, 2019, 4405-4410 4410  
  

www.etasr.com Jomaa et al.: Greenhouse Modeling, Validation and Climate Control based on Fuzzy Logic 

 

Xo [kgH2O kgair
-1
] Outside absolute humidity 

Xsat [kgH2O kgair
-1
] Absolute saturation humidity 

Zref [m] Reference depth 

α [°] Opening window angle  

∆ [kPa
  
°C

-1
] Slope of water vapor saturation 

γ [kPa
  
°C

-1
] Psychometric constant 

λ [J Kg
-1
] Latent heat of vaporization 

ρ [Kgair m
-3
] Air density [1.25 Kgair m

-3
] 

τ Transmission coefficient of the greenhouse 

REFERENCES 

[1] L. M. Mortensen, “Growth responses of some greenhouse plants to 
environment: III. Design and function of a growth chamber prototype”, 

Scientia Horticulturae, Vol. 16, No. 1, pp. 57-63, 1982  

[2] L. Meihui, D. Shangfeng, C. Lijun, H. Yaofeng, “Greenhouse Multi-

Variables Control by Using Feedback Linearization Decoupling 
Method”, Chinese Automation Congress, Jinan, China, Octomber 20-22, 

2017 

[3] M. Abbes, A. Farhat, A. Mami, G. D. Tanguy, “Pseudo bond graph 
model of coupled heat and mass transfers in a plastic tunnel 

greenhouse”, Simulation Modelling Practice and Theory, Vol. 18, No. 9, 
pp. 1327-1341, 2010  

[4] G. D. Pasgianos, K. G. Arvanitis, P. Polycarpou, N. Sigrimis, “A 

nonlinear feedback technique for greenhouse environmental control”, 
Computer and Electronics Agriculture, Vol. 40, No. 1-3, pp. 153-177, 

2003 

[5] C. J. Taylor, P. C. Young, A. Chotai, A. R. Mcleod, A. R. Glasock, 
“Modelling and proportional-integral-plus control design for free air 

carbon dioxide enrichment systems”, Journal of Agricultural 
Engineering Research, Vol. 75, No. 4, pp. 365-374, 2000 

[6] E. J. V. Henten, J. Bontsema, “Time-scale decomposition of an optimal 

control problem in greenhouse climate management”, Control 
Engineering Practice, Vol. 17, No. 1, pp. 88-96, 2009 

[7] X. Luan, P. Shi, F. Liu, “Robust adaptive control for greenhouse climate 

using neural networks”, International Journal of Robust and Nonlinear 
Control, Vol. 21, No. 7, pp. 815–826, 2011 

[8] R. Linker, I. Seginer, P. O. Gutman, “Optimal CO2 control in a 
greenhouse modeled with neural networks”, Computers and Electronics 

in Agriculture, Vol. 19, No. 3, pp. 289–310, 1998 

[9] P. M. Ferreira, E. A. Faria, A. E. Ruano, “Neural network models in 
greenhouse air temperature prediction”, Neurocomputing, Vol. 43, No. 

1-4, pp. 51-75, 2002 

[10] M. Azaza, K. Echaieb, F. Tadeo, E. Fabrizio, A. Iqbal, A. Mami, “Fuzzy 
decoupling control of greenhouse climate”, Arabian Journal for Science 

and Engineering, Vol. 40, No. 9, pp. 2805-2812, 2015 

[11] P. Salgado, J. B. Cunha, “Greenhouse climate hierarchical fuzzy 
modelling”, Control Engineering Practice, Vol. 13, No. 5, pp. 613–628, 

2005 

[12] M. Nachidi, F. Rodriguez, F. Tadeo, J. L. Guzman, “Takagi-sugeno 
control of nocturnal temperature in greenhouses using air heating”, ISA 

Transactions, Vol. 50, No. 2, pp. 315-320, 2011 

[13] M. Y. E. Ghoumari, H. J. Tantau, J. Serrano, “Non-linear constrained 
MPC: Real-time implementation of greenhouse air temperature control”, 

Computers and Electronics in Agriculture, Vol. 49, No. 3, pp. 345–356, 
2005 

[14] J. P. Coelho, P. B. de Moura Oliveira, J. B. Cunha, “Greenhouse air 
temperature predictive control using the particle swarm optimisation 

algorithm”, Computers and Electronics in Agriculture, Vol. 49, No. 3, 
pp. 330–344, 2005 

[15] Q. Zou, J. Ji, S. Zhang, M. Shi, Y. Luo, “Model Predictive Control 

Based on Particle Swarm Optimization of Greenhouse Climate for 
Saving Energy Consumption”, World Automation Congress, Kobe, 

Japan, September 19-23, 2010 

[16] M. Jomaa, F. Tadeo, A. Mami, “Modelling and Control of Greenhouses 
Powered by a Renewable Energy System”, 18th International 

Conference on Sciences and Techniques of Automatic Control and 
Computer Engineering, Monastir, Tunisia, December 21-23, 2017 

[17] X. Blasco, M. Martinez, J. M. Herrero, C. Ramos, J. Sanchis, “Model-
based predictive control of greenhouse climate for reducing energy and 

water consumption”, Computers and Electronics in Agriculture, Vol. 55, 
No. 1, pp. 49-70, 2007 

[18] K. A. Joudi, A. A. Farhan, “A dynamic model and an experimental study 

for the internal air soil temperatures in an innovative greenhouse”, 
Energy Conversion and Management, Vol. 91, pp. 76-82, 2015 

[19] R. C. Miranda, E. V. Ramos, R. R. Peniche-Vera, G. Herrera-Ruiz, 

“Fuzzy greenhouse climate control system based on a field 
programmable gate array”, Biosystems Engineering, Vol. 94, No. 2, pp. 

165-177, 2006