Optimization of Dynamical Systems

4th International
February Conference
1-5, 2016. on Advances
NIT Tiruchirappalli, in Control and
India
4th International Conference on Advances in Control
Available onlineand
at www.sciencedirect.com
Optimization
4th of Dynamical
International Conference Systems
on Advances in Control and
Optimization of Dynamical Systems
February 1-5, of
Optimization 2016. NIT Tiruchirappalli,
Dynamical Systems India
February 1-5, 2016. NIT Tiruchirappalli, India
February 1-5, 2016. NIT Tiruchirappalli, India
ScienceDirect
Fuzzy Based Temperature
IFAC-PapersOnLine Control
49-1 (2016) 549–554 of Greenhouse
Fuzzy
Fuzzy Based Temperature Control of Greenhouse
Fuzzy Based
Based Temperature
Temperature Control
ControlN**
Revathi S*, Sivakumaran of
of Greenhouse
Greenhouse

Revathi
Revathi S*,
S*, Sivakumaran
Sivakumaran N**
* Department of Instrumentation and ControlN**
Revathi S*, Sivakumaran

Engineering (e-mail:
N**
** Department revathisoundiran@gmail.com)

of Instrumentation
 and Control Engineering (e-mail:
*** Department
Departmentof
Department ofInstrumentation
of Instrumentation
Instrumentation and
andControl
and ControlEngineering
Control
revathisoundiran@gmail.com) Engineering(e-mail:
Engineering (e-mail:
(e-mail:
revathisoundiran@gmail.com)
nsk@nitt.edu)
** revathisoundiran@gmail.com)
** Department
Department
National
of
ofof
Institute
Instrumentation
Instrumentation
Technology
and
and Control
Control Engineering
Engineering
Tiruchirappalli,
(e-mail:
(e-mail:
Tamil Nadu, India.
** Department of Instrumentation and Control Engineering
nsk@nitt.edu) (e-mail:
nsk@nitt.edu)
nsk@nitt.edu)
National Institute of
National of Technology Tiruchirappalli, Tamil Nadu,
Nadu, India.
National Institute
Institute of Technology
Technology Tiruchirappalli,
Tiruchirappalli, Tamil
Tamil Nadu, India.
India.
Abstract: These Greenhouses should provide a controlled environment for plant production with sufficient
sunlight, temperature
Abstract: These and humidity.
Greenhouses should Better growing conditions are achieved inproduction
greenhouses mainly by
Abstract:
maintaining These
a Greenhouses
higher internal should provide
ambient provide
as
aa controlled
comparedcontrolled
with
environment
environment
external
for
for plant
plant
temperature. production
Thus the
with
with sufficient
greenhousesufficient
heater
Abstract:temperature
sunlight, These Greenhouses
and should Better
humidity. providegrowing
a controlled environment
conditions are for plantinproduction
achieved greenhouseswithmainly
sufficient
by
sunlight,
sunlight, temperature
requirements depend
temperature and
upon humidity.
the amount
and humidity. Better
of
Better growing
heat loss
growing conditions
from the are
structure.
conditions are achieved
The in
proposed
achieved Thus greenhouses
scheme
in greenhouses mainly
measures
mainly by
the
by
maintaining
maintaining a
a higher
higher internal
internal ambient
ambient as
as compared
compared with
with external
external temperature.
temperature. Thus the
the greenhouse
greenhouse heater
heater
on-line sequential
maintaining
requirements adepend data
higher of temperature
internal
upon theambient
amount from
as
of theloss
compared
heat greenhouse and the
with external
from the heating
temperature.
structure. The power
Thus is
proposed recursively
the greenhouse
scheme updated
heater
measures the
requirements
requirementsthe depend
based onsequential
energy
depend upon
upon the
theofamount
balance of heat
an elementary
amount of heat loss from
fromofthe
volume
loss structure.
thegreenhouse
structure. The
air by
The proposed scheme
using intelligent
proposed scheme measures the
controllers.
measures the
on-line
on-line sequential data
data of
of temperature
temperature from
from the
the greenhouse
greenhouse and
and the
the heating
heating power
power is
is recursively
recursively updated
updated
Simulation
on-line
based on results
sequential
the energyof the
data ofgreenhouse
temperature
balance of dynamics
from
an elementarythe illustrate the
greenhouse
volume effectiveness
and the
of greenhouse of
heating
air bythe
powerproposed
using is scheme
recursively
intelligent without
updated
controllers.
based on
the exact
based the energy
on mathematical
the energy balance
model
balance of
of an
of theelementary
an plant.
elementary volume
volume of ofthegreenhouse
greenhouse air by
air of using
bythe
using intelligent
intelligent controllers.
controllers.
Simulation
Simulation results
results of
of the
the greenhouse
greenhouse dynamics
dynamics illustrate
illustrate the effectiveness
effectiveness of the proposed
proposed scheme
scheme without
without
Simulation
Keywords:
the
© exact
2016, results
mathematical
IFAC of
Greenhouse, the greenhouse
fuzzy
model
(International logic
of the
Federation dynamics
controller,
plant.
of illustrate
heat
Automatic loss, the
Control) effectiveness
temperature
Hosting by of
control, the
Elsevier proposed
heating
Ltd. power,
All scheme
rights PSO. without
reserved.
the exact mathematical model of the
the exact mathematical model of the plant. plant.
Keywords:
Keywords: Greenhouse,
Greenhouse, fuzzy
fuzzy logic
logic controller,
controller, heat
heat loss,
loss, temperature
temperature control,
control, heating
heating power,
power, PSO.
PSO.
Keywords: Greenhouse, fuzzy logic controller, heat loss, temperature control, heating power, PSO.
upon the amount of heat loss from the structure. Heat loss from
1. INTRODUCTION 

 aupon greenhouse
the amount usually
of heat occurs
loss frombytheallstructure.
three modesHeat of from
loss heat
Greenhouse cultivation 1. INTRODUCTION
aims basically to protect the upon
upon the
transfer:
the amount
conduction,
amount of
of heat
heat loss
convection
loss from
from the
and
the structure.
radiation.
structure. Heat
Usually
Heat loss
loss from
many
from
1.
1. INTRODUCTION
INTRODUCTION aa greenhouse usually occurs by all three modes of heat
plantations from bad weather and becomes intorecent years athe types greenhouse
of heat
atransfer:
greenhouse usually
exchange occurs
occur
usuallyconvection by
by all
occurssimultaneously. three
three modes
allradiation. modes of heat
of many
heat
Greenhouse cultivation aims basically protect transfer: conduction,
conduction, convection and
and radiation. Usually
Usually many
Greenhouse
mean to achieve
Greenhouse cultivation
controlled
cultivation aims
aims basically
agricultural to
basically inproduction.
to protect
protect Thethe transfer: conduction, convection and radiation. Usually many
plantations
plantations from
from bad
bad weather
weather and
and becomes
becomes in recent
recent years
years aathe types
types of heat exchange occur simultaneously.
of heat exchange occur simultaneously.
control
plantationsof the climatic
from bad weather andenvironment in greenhouses
becomes inproduction.
recent yearsThe has
a 2.1 Conduction
types of heat exchange occur simultaneously.
mean
mean to achieve
achieve controlled
to considerable controlled agricultural
agricultural production. The
received
mean
control toofachieve
the climaticattention
controlled in these in
agricultural
environment lastgreenhouses
years. The main
production. The
has 2.1 Conduction
control
reasons of this
for the increasing
climatic environment
interest are in greenhouses
related to the The has
following 2.1
2.1 Conduction
Heat is conducted either through a substance or between
Conduction
control
received of the climatic
considerable environment
attention in these in
last greenhouses
years. has
main
received
agronomic considerable
and financial attention in
objectives: these last
(a)related
tolast years.
extend The
theThe main
growing objects by direct physical contact. aThe rate of or conduction
received
reasons considerable
for this increasingattention in are
interest these years.
to the main
following Heat is conducted either through substance between
reasons
season for
and this increasing
the financial interest
potential objectives:
yield; (b)areto related
manage to the following Heat
betweenis conducted
two objectseither through
depends on aa the
substance
area, or
path between
length,
reasons
agronomic for and
this increasing interest are (a)related
to extendtothe
the
theclimate
following
growing in Heat
objectsis conducted
by direct either
physical through
contact. Thesubstance
rate of or between
conduction
agronomic
order to and
reach financial
higher objectives:
standards of (a)
quality; to extend
(c) to the
developgrowing
low- objects
objects by
temperature
by direct
direct physical
difference
physical and contact.
physical
contact. The
The rate of
properties
rate of conduction
of
conduction the
agronomic
season and and
the financial
potential objectives:
yield; (b) (a)manage
to to extend the the growing
climate in between
between two
two objects
objects depends
depends on
on the
the area,
area, path
path length,
length,
season
cost
season and
production
and the potential
systems,
thehigher
potential yield; (b)
compatible
yield; to
to manage
(b)quality; with
manage the
the
the climate
scarcity
climate in
of
in substance(s)
between
temperature two(such as density).
objects
difference depends
and Heat
on transfer
physical by conduction
the properties
area, path of length, is
the
order
order to
to reach
reach higher standards
standards of
of quality; (c)
(c) to
to develop
develop low-
low- temperature difference and physical properties of heatthe
resources and higher
thesystems,
low investment capacity (c)of to growers. The most easily
temperature
substance(s) reduced
difference
(such by replacing
and Heat
as density). a material that
physical properties conducts of the
order
cost
cost
to reach
production
production
standards of quality;
compatible with the develop
scarcity low- of substance(s)
rapidly a(such
poor as
withreduced density).
thermal Heat transfer
conductor transfer by
by conduction
(insulator) conduction
or by placing
is
is
main
cost objective
production
resources and ofsystems,
the the
systems,
low
compatible
climate control
compatible
investment
with
problem
with of
capacity
the
the scarcity
isgrowers.
to maintain
scarcity The
of
of substance(s)
most easily (such asbydensity).
replacing Heat
a transfer
material by conducts
that conduction heatis
resources and the low investment capacity of growers. The most
an
most easily
insulator
easily reduced
in the heat
reduced by replacing
by flow a material
path.a material
replacing that conducts
that conducts heat
heat
the temperature
resources
main and
objective the inside
low
of the greenhouse
investment
climate control environment
capacity
problem of growers.
is within
The
to maintain rapidly with a poor thermal conductor (insulator) or by placing
main objective rapidly with aa poor thermal conductor (insulator)
(insulator) or or by
by placing
suitable
main
the ranges. of
objective
temperature The
of the
the climate
difficulty
climate
inside
control
resides
control in
greenhouse
problem
the
problem is
is to
complexity
environment to maintain
of the
maintain
within rapidly
an with in
insulator poor
the thermal
heat flow conductor
path. placing
the temperature inside greenhouse environment within an
an insulator
2.2insulator
Convection in the heat flow
in the heat flow path. path.
phenomena
the temperature
suitable ranges. which
The conditions
inside greenhouse
difficulty residesthe in ideal
environment environment,
the complexity within
of the
suitable
suitable ranges.
essentially The
dictated
ranges. Theby difficulty resides
the day/night
difficulty resides in
in the
cycle,the complexity
the growth
complexity of
of the
season,the 2.2 Convection
phenomena
phenomena which
which conditions
conditions the
the ideal
ideal environment,
environment, 2.2 Convection
Convection heat transfer is the physical movement of a warm
the local
phenomena climate,
whichand the nature
conditions of the
the culture.
ideal (Bakker J
environment, C et 2.2 Convection
essentially
essentially dictated
dictated by
by the
the day/night
day/night cycle,
cycle, the
the growth
growth season,
season, gas or liquidheat to atransfer
colder is location. Heat movement
losses by convection
al.,1995).
essentially
the local dictated
climate, and by the day/night
the nature of thecycle, the growth
culture. (Bakker season,
J C et Convection
Convection heat transfer is the
the physical
physical movement of
of a warm
the
the local
local climate,
climate, and
and the
the nature
nature ofof the
the culture.
culture. (Bakker
(Bakker JJ C C etet
inside
Convection
gas or the
liquid greenhouse
heatto transfer
a colder is occur through
the physical
location. Heat movement
losses by of aa warm
ventilation and
warm
convection
al.,1995). gas or liquid
infiltration (fansto aandcolder
air location. Heat losses by convection
leaks).
al.,1995).
al.,1995). 2. HEATING gas or liquid
inside the to a colder location.
greenhouse occur Heat
through losses by convection
ventilation and
inside
inside the greenhouse
the(fans
greenhouse occur
occur through
through ventilation
ventilation and and
infiltration
infiltration (fans and air leaks).
Many greenhouses must 2.
2. HEATING
be heated for year-round crop 2.3 Radiation
infiltration (fans and and air
air leaks).
leaks).
production. A good 2. HEATING
HEATING
heating system is one of the most
Many greenhouses must be heated for year-round crop 2.3
2.3 Radiation
Radiation
Many
Many greenhouses
important steps
greenhouses to must
successful
must be heated
plant
be system for year-round
production.
heated for crop
Anytheheating
year-round crop Radiation
2.3 Radiationheat transfer occurs between two bodies without
production.
production. A
A good
good heating
heating system is
is one
one of
of the most
most
system that
production. provides
A good uniformplant
heating temperature
system is onecontrol
of thewithout
most direct
Radiation contact
heat ortransfer
the needoccurs
for a medium
between such
two asbodies
air. Like light,
without
important steps to successful production. Any heating Radiation heat transfer
important
releasing
important steps
material
steps to successful
toharmful
successful plant
to the
plant production.
plants Any
is acceptable.
production. heating
Any Suitable
heating heat radiation
Radiation
direct heatorfollows
contact transfer
the needa occurs
straight
occurs
for a
between
line and
between
medium
two
two
such
bodies
isaseither
bodies
air. Like
without
reflected,
without
light,
system that provides uniform temperature control without direct contact
system
energy
system
releasing
that
sources
that provides
include
provides
material
uniform
natural
uniform
harmful to the
temperature
gas, LP gas,
temperature
plants is
control
fuel oil,
control
acceptable.
without
wood and
without
Suitable direct
heat contactoror
transmitted
radiation
the
the need
absorbed
orfollows needa
for
upon
for aa medium
straight striking
medium
line
such
an
such
and is
as air.
air. Like
object.
aseither Like light,
Radiant
light,
reflected,
releasing material
electricity. The harmful
cost and to the
availabilityplants
of is acceptable.
these sources Suitable
will vary heat
energy radiation
must be follows
absorbed a straight
to be line and tois heat.
converted eitherThe reflected,
rate of
releasing
energy material harmful
sources include to the
natural gas, plants is
LP gas, acceptable.
fuel oil, Suitable
wood and heat radiation
transmitted or follows
absorbed a straight
upon line andan
striking is either
object. reflected,
Radiant
energy transmitted or absorbed upon striking an object. Radiant
energy sources
somewhat from
sources
electricity. The
include
areanatural
one and
include
cost natural gas,
gas, LP
to another.
availability LP
of
gas,
gas, fuel
Convenience,
these fuel oil,
sources
wood
wood
will
and
oil,investment
and
vary
radiation
transmitted
energy heat
must or
betransfer varies
absorbed to upon
be with the area
striking an
converted to of an
object.
heat. object,
The Radiant
rateand
of
electricity. The cost and availability of these sources will vary energy must
temperature be
and absorbed
surface to be converted to heat. The rate of
and operating
electricity.
somewhat costs
The cost
from one are
and
area all further
availability
to another. considerations.
of these sources
Convenience, Savings in
will vary
investment energy
radiationmust
heat be absorbed
transfer tocharacteristics
varies bewith
converted
the areaofof
to the
heat.
an two
The bodies
object, rateand
of
somewhat
labour couldfrom one area
justify a to another.
more Convenience,
expensive heating investment
system with radiation
involved.
radiation heat
heatandtransfer
transfer varies
varies with the area
with the areaofoftheof an object,
an two
object, and
and
somewhat
and fromcosts
operating one area
are to further
all another.considerations.
Convenience, investment
Savings in temperature surface characteristics bodies
and operating costs are all temperature and
and surface characteristics of
of the
the two
two bodies
automatic
and controls.
operating
labour could justify are
a all further
costs Greenhouse
more further considerations.
heating requirements
considerations.
expensive heating
Savings
Savings
system
in
depend
in
with
2.4 Factors Affecting
temperature
involved. Heating
surface Loss
characteristics bodies
labour could justify a more expensive heating system with involved.
involved.
labour could
automatic justifyGreenhouse
controls. a more expensive
heating heating system
requirements with
depend 2.4 Factors Affecting Heating Loss
automatic 2.4
2.4 Factors
Factors Affecting
Affecting Heating
Heating Loss
automatic controls.
controls. Greenhouse
Copyright © 2016 IFAC
Greenhouse heating
heating requirements
requirements depend
depend 549 Loss
2405-8963 © 2016, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.
Peer review©under
Copyright 2016 responsibility
IFAC of International Federation of Automatic
549Control.
Copyright © 2016 IFAC
10.1016/j.ifacol.2016.03.112 549
Copyright © 2016 IFAC 549
IFAC ACODS 2016
550 Revathi S et al. / IFAC-PapersOnLine 49-1 (2016) 549–554
February 1-5, 2016. NIT Tiruchirappalli, India

At an Augusta location and an average daily minimum

Heat loss by air infiltration depends on the age, condition and January temperature of 33.6 degrees F, the design
type of greenhouse. Older greenhouses or those in poor temperature would be about 18.6 degrees F, so use 20
condition generally have cracks around doors or holes in degrees F. This requires a 45-degree F rise above design
covering material through which large amounts of cold air may temperature; and, with glass, the R-value will be 0.91.
enter. Greenhouses covered with large sheets of glazing
Conduction Heat Loss, Q C = Area x ΔT/R (1)
materials, large sheets of fibre glass or a single or double layer Air Infiltration Losses, Q A = 0.02 x Volume x C x ΔT (2)
of rigid or flexible plastic have less infiltration (Fig. 1). Solar Perimeter Heat Loss, Q P = P x L x (ΔT) (3)
radiation enters a greenhouse and is absorbed by plants, soil Total Heat Loss, Q T = Q C + Q A + Q P (4)
and greenhouse fixtures. The warm objects then re-radiate this
Where
energy outward. The amount of radiant heat loss depends on
the type of glazing, ambient temperature and amount of cloud Q = Heat loss, BTU/hr
cover. (Seginer,et al.; 1992) A = Area of greenhouse surface, sq. ft
R = Resistance to heat flow (a characteristic of the material)
V = Greenhouse volume, cu. ft
C = Number of air exchanges per hour
P = Perimeter heat loss coefficient, BTU/ft °F hr
L = Distance around perimeter
The values are listed below in the table 2.
Table 2.Values used for the greenhouse considered
R(glass) 0.91
V 30,928 cu ft
C for new 0.75 to 1
construction, glass
or fiberglass
C for Old 1 to 2
Construction glass,
Fig. 1. Energy losses and gains in a greenhouse good maintenance
C for Old 2 to 4
Construction glass,
3. CALCULATION OF MINIMUM DESIGN poor condition
TEMPERATURES Perimeter for 0.8 BTU/ft °F hr
A good outside temperature to use in heater design calculations Uninsulated P
(to select heater size) can be found by subtracting 15 degrees Perimeter for 0.4 BTU/ft °F hr
F from the average daily minimum temperature. Another insulated P
requirement the heater must meet is to provide enough heat to
prevent plants from freezing during periods of extremely low 4. TEMPERATURE CONTROL
temperatures. (as cited in
www.cps.gov.on.ca/english/plans/E6000/6701/M-6701L.pdf) Temperature is the main climatic variable affecting the crop
For example consider the table 1. growth, and thus, this variable has been traditionally
controlled in the greenhouses. Based on the energy balance
Table 1.Climatic Conditions in Georgia (1948-2004)
of an elementary volume of greenhouse air, the relation can
Location Minimum Temperature Average Daily Minimum be given by eq. 5. (Arvanitis, 1999).
°F and (Year Occurring) January Temperatures
(°F) 𝑑𝑑𝑇𝑇𝐺𝐺
𝑄𝑄𝐻𝐻 = 𝐶𝐶𝑞𝑞 ( ) − (𝐾𝐾𝑜𝑜𝑜𝑜𝑜𝑜,𝑎𝑎𝑎𝑎𝑎𝑎 [𝑇𝑇𝑜𝑜𝑜𝑜𝑜𝑜 −𝑇𝑇𝐺𝐺 ]) (5)
𝑑𝑑𝑑𝑑
Atlanta -8 (1985) 33.6

Athens -4 (1985) 33.2

Where 𝑇𝑇𝐺𝐺 is the greenhouse temperature, 𝐶𝐶𝑞𝑞 is the greenhouse
thermal capacity, 𝐾𝐾𝑜𝑜𝑜𝑜𝑜𝑜,𝑎𝑎𝑎𝑎𝑎𝑎 is the heat loss coefficient from
Augusta -1 (1985) 33.6
greenhouse air to outside air, 𝑇𝑇𝑜𝑜𝑜𝑜𝑜𝑜 is the outside temperature
Columbus -2 (1985) 36.4
and 𝑄𝑄𝐻𝐻 is the heating power.
Macon -6 (1985) 35.8
Rome -9 (1985) 30.5 The scheme of temperature control inside the greenhouse is
Savannah 3 (1985) 39.0 shown in (Fig. 2).
Tifton 0 (1985) 38.0
Valdosta 9 (1981) 38.6

550
 IFAC ACODS 2016
Revathi S et al. / IFAC-PapersOnLine 49-1 (2016) 549–554 551
February 1-5, 2016. NIT Tiruchirappalli, India
Fig. 2. Greenhouse heating system and control
The outside temperature and the present inside temperature of
the greenhouse are the inputs from the sensor. The error and
the change in error are computed which are the inputs to the
fuzzy logic controller. The output is the heating power
required to maintain the greenhouse at an optimum
temperature.
5. FUZZY CONTROL
In the last decade, the fuzzy logic gained interest in the
scientific community and fuzzy knowledge-based systems are
one of the most successful applications of fuzzy sets and fuzzy
logic methods. This is mainly due to the flexibility and
simplicity by which knowledge can be expressed using fuzzy
rules as well as to the theoretical developments in this field.
The aim advantage of fuzzy control is the possibility of
implementing human expert knowledge in the form of
linguistic if – then rules. The design of a fuzzy controller
begins with the choice of linguistic variables, the process state,
the input and the output variables. The next step is the choice
of the set of linguistic rules and the kind of fuzzy reasoning
process. Once the rules are setup, after the inference, the fuzzy
set and the crisp output value have to be generated; a
defuzzification strategy has to be established too. (Horiuchi J,
2002)
Fig. 3. Block diagram of a fuzzy controller
The block diagram of a fuzzy controller is shown in fig 3. It is
composed of four principal modules: The fuzzification
interface performs the transformation of crisp inputs into fuzzy
sets. The knowledge base supplies the fuzzification module,
the interface engine, and the defuzzification interface with
necessary information for their proper functioning. The
decision making unit or interface engine computes the
meaning of the set linguistic rules. The defuzzification
interface transforms the union of fuzzy sets (individual
551
contributions of each rule in the rule base) into a crisp output.
(Chuen-Lee, C., 1990)
5.1 Optimized fuzzy logic control using PSO
In order to increase the degree of automation of the fuzzy
control system, a “near-optimal” controller synthesis strategy
has been developed using PSO technique. An optimization
method that finds the optimal solution using a population of
particles is PSO algorithm. Each swarm of PSO is a solution
in the solution space. The algorithm can be explained as
follows. (Eberhart, R. 1995)( Lafont 2002).
• Each individual particle i has the following properties: A
current position in search space Xid , a current velocity pid and
a personal best position in search space Pid .
• The personal best position pid corresponds to the position
in search space where particle i presents the smallest error as
determined by the objective function 𝑓𝑓 assuming a
minimization task.
• The global best position marked by represents the
position yielding the lowest error amongst all the pgd .
During the iteration every particle in the swarm is updated
using the following two equations:
Vid (t + 1) = w. Vid (t) + c1 . r1 (Pid − Xid (t))c2 . r2 (Pgd − Xid (t))
Xid (t + 1) = Xid (t) + Vid (t + 1) (6)
Where Vid (t + 1) and Vid (t)are the updated and current
particles velocities, respectively, Xid (t + 1) and Xid (t) are
the updated and current particle position, respectively, 𝑐𝑐1 and
𝑐𝑐2 are two positive constants and r1 and r2 are normalized unit
random numbers within the range [0, l] and w is the inertia
weight.
The approach of using a PSO for MF tuning in FLC is shown
in (Fig. 4).
Fig. 4. Flowchart for Optimizing FLC using PSO
IFAC ACODS 2016
552 Revathi S et al. / IFAC-PapersOnLine 49-1 (2016) 549–554
February 1-5, 2016. NIT Tiruchirappalli, India

In this process, each particle is shaped to represent the MF

parameters of the FLC’s inputs and outputs. The cost function
used here is root mean square error. The PSO parameters are
listed below in the table 3.

Table 3. PSO parameters

PSO parameters values

Population Size 25

Personal Learning Coefficient 1

Global Learning Coefficient 2

Fig. 7. Membership function for ce
Inertia Weight 1
The labels used for linguistic variables ‘out’ are ‘range1’,
Maximum Number of Iterations 1000 ‘range2’, ‘range3’, ‘range4’, ‘range5’, ‘range6’. Triangular
and trapezoidal membership functions are used.
6. SIMULATION
For the simulation purpose the weather data of Georgia is
taken for the month of January 2015 and used as input external
temperature outside the greenhouse.

Fig. 8. Membership function for the output

Fig. 5. Temperature Graph of Georgia (January 2015)
(www.accuweather.com/en/us/ga/georgia.weather)
6.1 Linguistic variables, values and membership functions of
basic fuzzy logic control
The minimum temperature pattern recorded in Georgia in
January 2015 is shown in (Fig. 9). It is compared with against
the greenhouse temperature. The error and change in error are
given as input to Fuzzy controller.

The linguistic variables are error e, change in error ce and the

output u. The membership functions are shown in (fig 6, 7, 8).
The labels used for linguistic variables ‘e’ are ‘extreme –ve’,
’too-ve’, ’low-ve’, ‘small-ve’, ‘zero’, ’small+ve’ and ‘+ve’.

Fig. 9. Minimum temperature of Georgia in January 2015

Fig. 6. Membership function for e Fuzzy controller evaluates the heating power that is to be
applied to the greenhouse using 21 rules and it is listed in table.
The labels used for linguistic variables ‘ce’ are ‘low’, ‘med’ 4 and the pattern is shown in (Fig. 10). These rules are based
and ‘high’. on the knowledge of heating requirement of greenhouse
mentioned in section 3 of the paper.

552
 IFAC ACODS 2016
Revathi S et al. / IFAC-PapersOnLine 49-1 (2016) 549–554 553
February 1-5, 2016. NIT Tiruchirappalli, India

Table 4. Fuzzy rules functions which are smooth and non-zero at all points. The
input membership functions error ‘e’ and change in error ‘ce’
Ce/e extr too- low- sma zero sma +ve are shown in (Fig. 12, 13).
eme ve ve ll- ll+v
–ve ve’ e’
low Ran Ran Ran Ran Ran Ran Ran
ge6 ge6 ge5 ge4 ge3 ge2 ge1

med Ran Ran Ran Ran Ran Ran Ran

ge6 ge6 ge5 ge4 ge3 ge2 ge1

high Ran Ran Ran Ran Ran Ran Ran

ge6 ge5 ge4 ge3 ge2 ge1 ge1

The output of a process can be logical union of two or more

Fig. 12. Optimized membership function of e
membership functions defined on the universe of discourse the
output value. Centroid membership function is used for
defuzzification process. The output variable is the heating
temperature value applied to the greenhouse and it is shown in
(fig. 10).

Fig. 13. Optimized membership function of ce

FIS is Takagi-Sugeno type fuzzy controller so the

Fig. 10. Heating Power pattern applied to the greenhouse for
January 2015.
(Fig. 11) shows the controlled temperature using basic FLC
inside the greenhouse throughout the month of January 2015.
defuzzification membership functions are Singleton. Singleton
membership functions are simple and use less computing
resources. Also the using PSO, the rules are optimized and 7-
rule FLC provide the control strategy.
450

400

350
Heating pattern in KW

300

250

200

150
0 5 10 15 20 25 30
Days of the Month(Jan 2015)
Fig. 11. FLC Controlled Temperature response inside
glasshouse Fig. 14. Optimized Heating Power pattern applied to the
greenhouse
6.2 Optimized membership functions of fuzzy logic control
The controller output is also optimized and also the
During optimization, Gaussian membership functions instead computation memory required and associated time is also
of triangular have been adopted in order have membership reduced.

553
IFAC ACODS 2016
February 1-5, 2016. NIT Tiruchirappalli, India Revathi S et al. / IFAC-PapersOnLine 49-1 (2016) 549–554
554

Optimized Fuzzy controlled Temperature response inside glasshouse

19.5
Horiuchi J (2002). Fuzzy modelling and control of biological
19 processes. Journal of Bioscience and Bioengineering,
94(6), 574–578.
18.5

Lafont. F and Balmat .J. -F. (2002). Optimized fuzzy control

Temperature in Degree celcius

18

of a greenhouse.Fuzzy Sets and Systems, 28, 47-59.

17.5

17
Seginer, I., McClendon, R.W., (1992). Methods for optimal
control of the greenhouse environment, Trans. ASAE
16.5
35, 1299-1307.
16

www.accuweather.com/en/us/ga/georgia.weather
15.5

15
www.cps.gov.on.ca/english/plans/E6000/6701/M-6701L.pdf
0 5 10 15 20 25 30
Days of month (Jan 2015)

Fig. 15. Optimized FLC Controlled Temperature response

inside glasshouse

From (Fig 10, 11, 14 and 15) it is inferred that better

improvement in temperature inside greenhouse is achieved
with similar input energy pattern. Also from the heating
pattern of both basic fuzzy and optimised fuzzy control, it
shows PSO-FLC has more acceptable desired results rather
than conventional FLC in control of the greenhouse
temperature. Also PSO-FLC controller gives robustness
improvement and very good results in compare with the
conventional FLC controller.

7. CONCLUSION
Due to the greenhouse complex dynamics of the envelopment
physics, the fuzzy control represents a useful tool to be applied
to this type of processes. In this paper, greenhouse temperature
is controlled using basic fuzzy and optimized fuzzy through
PSO. It is inferred that the optimized membership functions
(MFs) provided better performance than a fuzzy model for the
same system, when the MFs were heuristically defined. Thus
throughout the month including severe weather conditions the
intelligent controllers are able to maintain temperature
between 15 and 20 degree Celsius without exact mathematical
model of the greenhouse. Thereby damage of the crops and
production losses are greatly reduced.
REFERENCES
Arvanitis, K.G.(1996). Adaptive LQ regulation by multirate-
output controllers, Found. Comput. Decis., Sci. 21,
183-213.

Bakker J C; Bot G P A; Challa H; De Braak N J V (1995).

Greenhouse Climate Control: An Integrated
Approach. Wageningen, The Netherlands.

Chuen-Lee, C. (1990b). Fuzzy Logic in Control systems:

Fuzzy logic controller, Part II. IEEE Transactions on
Systems Man and Cybernetics, 20, 419–435.

Eberhart, R. and Kennedy, I (1995).A new optimizer using

particle swarm theory. Symposium on Micro
Machine and Human Science, 39-43.

554