10-A Fuzzy Inference System For Air Quality in Using MATLAB, Chennai, India.

J. Environ. Res. Develop.
Journal of Environmental Research And Development Vol. 7 No. 1A, July-September 2012
484
*Author for correspondence
A FUZZY INFERENCE SYSTEM FOR AIR QUALITY
IN USING MATLAB, CHENNAI, INDIA
Kumaravel R.*
1
and Vallinayagam V.
1. SRM University, Career Development Centre, Kattankulathur, Chennai (INDIA)
2. Department of Mathematics, St.Josephs College of Engineering, Chennai (INDIA)
Received April 25, 2012 Accepted August 05, 2012
ABSTRACT
Air quality is an important issue of relevance in the context of present times. The proposed model is
designed to predict Air Quality Index (AQI) for Chennai in Tamil Nadu, India. A Fuzzy Inference
System (FIS) simplifies and speed up the computation of AQI as compared to the currently existing
standards. In this paper, the proposed model is compared with Indian Air Quality Index (IAQI) and it
is found that the designed model results in accurate prediction. The earths atmosphere is a mixture
of gases and particulate-phase substances. The most abundant of these, Nitrogen(N
2
) and Oxygen(O
2
),
comprise approximately 78% and 21% respectively, of atmosphere mass and volume. Air quality can
be defined qualitively. It is poor when cause a reduction in visibility, soil building surfaces and
damage materials, damage crops and other plants, cause adverse health effects. It is deemed good
when the sky appears clean and no adverse environmental effects are evident.
Key Words : Air Quality Index (AGI), Fuzzy logic, Fuzzy Inference System (FES), MATLAB
INTROUDCTION
The atmosphere that surrounds the plant earth is
one of the factors that makes our plant hospitable
towards life. The average human makes use of
about 30 pounds of air each day, using it to oxidize
food for energy and warmth. A number of trace
gases make up the remaining 1% average
concentration
1
in Table 1.
Table 1 : Percentage and volume with
different gases
Gases Percentage of volume
Nitrogen 78.03
Oxygen 20.99
Argon 0.94
Carbon di oxide 0.03
Balance* 0.01
*Includes hydrogen, helium , neon, krypton, xenon and vari-
ous pollutants
Contamination or pollution of the atmosphere
occurs as a consequences of natural process as
well as human activity ( the letter are called
anthropogenic). Anthropogenic air pollution as
been continues to be viewed as a serious
environmental and public health problem. Its
seriousness lies in the fact that elevated pollutant
levels are produced in environments where harm
to human health
2
and welfare is more likely.
The quality of air is our communities, country side
and even remote locations changes form hour to
hour, day to day and over longer time scales. Air
quality
3
can be defined qualitively. It is poor when
cause a reduction in visibility, soil building surfaces
and damage materials, damage crops and other
plants, cause adverse health effects. It is deemed
good when the sky appears clean and no adverse
environmental effects are evident.
The AQI is used to report daily air qualities in
metr opolitan areas for four pollut ions
O
3
,PM(PM
10
and PM
2.5
), CO, SO
2
and NO
2
.
Chennai, the fourth largest city in India, capital of
the state Tamil Nadu and is located on the eastern
sea- coast of the sub-continent at an latitude
3
of
1304" North and a longitude of 8015" East. It is
located at the center of an extensive network of
surface, sea and air transportation facilities
serving South India and is well connected to the
485
Fig. 2 (a) : Anna nagar map Fig. 2(b) : Kilpauk map
rest of India. The Chennai Metropolitan Area
(CMA) comprises of the city of Chennai and its
outlying urban and rural areas.
Chennai has a tropical climate with a mean annual
temperature of over 30
o
C and it varies from 20
o
C
to 45
o
C. Humidity ranges from 50% to 90%, with
the highest prevailing during monsoon. Cool sea
breezes blows almost throughout the year
providing relief from the normally high
temperature. The experiences an average rainfall
of about 110 cm , mostly during the north east
monsoon period between October and December.
The south west monsoon occurs between June
and September, during which period also the city
gets some rainfall. Rise in population and growth
in economic activity have led to increase in
pollution in Chennai. With good infrastructure
facilities, the city has become a major center for
commerce, industry and education. At the same
time, the civic amenities have not kept pace and
the unabated migration of floating population has
compounded the problem. The land use
regulations are not strictly followed as the
awareness among the public is lacking.
MATERIAL AND METHODS
In this paper, five locations are considered for
present study. Anna Nagar, Adyar, Kilpauk ,
Thiyagaraya Nagar, Vallalar Nagar in Chennai,
Tamil Nadu. These locations are shown in
Fig. 1 to Fig. 3.
Fig. 1(a) : Chennai map Fig. 1(b) : Adyar map
486
Fig. 3(a) : Thiyagaraya nagar map
Fig. 3(b) : Vallalar nagar map
Fuzzy logic
The fuzzy logic set was introduced in 1965 as
mathematical way to represent linguistic
vagueness(Zadeh)
4
. According to the fuzzy logic
concept, factors and criteria can be classified
without certain limits. Fuzzy logic is very useful
for addressing real-world problems, which usually
involve a degree of uncertainty. Fuzzy logic uses
variables like low, medium and high in place, of
true/false or yes/no variables. Fuzzy sets are
determined by membership functions. The
membership function of a fuzzy set is expressed
as SO
2
and membership degree of its fuzzy set is
determined as a number between 0 and 1.
The structure of a fuzzy rule based system is the
fuzzy algorithm the fundamental concepts of which
are derived is the fuzzy algorithm the fundamental
concepts of which are derived from fuzzy logic
5,4
.
The rules represent the relationship between the
inputs and the output of the system. The
generalized structure of a fuzzy system is
presented in (Fig. 4) Two commonly used
inference system are, i.e. Mamdani fuzzy model
and TSK fuzzy model. They use different types
of fuzzy reasoning and formulation of fuzzy if then
rules. Mamdani fuzzy model
6
is based on the
collections of if then rules with both fuzzy
antecedent and consequent parameters. Takagi
and sugeno proposed sugeno fuzzy model
7
, where
as sugeno and kang
8
built up a methodical
approach to generating fuzzy rules from a given
input-output data. These model are built with if-
then rules that have fuzzy antecedent and
functional consequent.
Fig. 4 : A Fuzzy Inference System for air quality index
487
Table 2 : Indian Air quality criteria classes (CPCB)
11
Requirement for air quality of class
S/N Parameters
G-Good M-Moderate P-Poor
a SO
2
0-80 70-300 250-400
b NO
2
0-80 60-180 170-500
c RSPM 0-200 180-260 250-400
d PM 0-400 350-550 500-900
Fuzzy system is comprised of four block viz:
Fuzzification, fuzzy rule base, fuzzy decision,
defuzzification
9
.
Fuzzification
is the operation of transforming a crisp set to a
fuzzy set. The operation translates crisp input or
measured values into linguistic concepts by using
suitable membership functions.
Fuzzy secision and rule base
Once the inputs are fuzzified, the corresponding
inputs fuzzy sets are passed to the fuzzy decision
that processes current inputs using the rules
retrieved from the rule base.
Defuzzification
At the output of the fuzzy inference there will
always be a fuzzy set that is obtained by the
composition of the fuzzy sets output by each of
the rules. In order to be used in the real world,
the fuzzy output needs to be interfaced to the crisp
domain by the defuzzifier by using suitable
membership functions.
Air Quality Index (AQI)
Air quality index is ranging from 1 to 300; a lower
number indicative of better Air quality.AQI rating
scale is as follows : 0-50 good Air quality(A),51-
150, Moderate Air quality(B),151-300:Unhealthy
Air quality(C)
10
. The parameters used in defining
AQI are SO
2
(sulphur di oxide),NO
2
(oxides of
nitrogen), RSPM (respirable dust particulate
matter, < 10 micron size),PM (particulate matter,
< 2.5 micron size). The criteria given in (Table 2)
to determine how healthy the air is on a given day.
Fuzzy Inference System (FIS)
In this paper, an attempt has been made to use
Fuzzy Inference System (FIS) model to predict
(or)estimate the Air Quality Index (AQI) given
SO
2
(sulphur di oxide),NO
2
(oxides of nitrogen),
RSPM (respirable dust particulate matter, < 10
micron size), PM (particulate matter, < 2.5 micron
size) as input parameters. Fig. 5 represents the
architecture of the Mamdani fuzzy expert system.
As maintained in the inputs SO
2
,NO
2
,RSPM,PM.
The output of the system are taken as IAQI.
Hence, it is a very helpful to estimate AQI to
design a fuzzy inference system. The methodology
for the development of the fuzzy Inference
System (FIS) based Air Quality index(AQI) model
involves the following steps.
1. Fuzzification of input and output variables.
2. Selection of membership functions for input
and output variables.
3. Determination of application rule base.
4. Defuzzification of IAQI.
Fuzzification of input and output variables
First, we select the system variables i.e. the fuzzy
system modeling is the identification of input and
output variables. The most important input
variables are SO
2
,NO
2
,RSPM,PM. Suppose
select of more number of inputs to the system
inputs to the system requires more number of rules
and hence the complexity increases. The input
and output are taken in the form of linguistic
format. For example, SO
2
=(Good, Moderate,
Poor), NO
2
=(Good, Moderat e, Poor),
RSPM=(Good, Moderate, Poor), PM=(Good,
Moderate, Poor). The output variable similarly
divided into Indian AQI=(A-Good, B-Moderate,
C-Unhealthy).
488
Fig. 5 : Determination of air quality index using Fuzzy logic
Selection of membership functions for input
and output variables
Linguistic values are expressed in the form of
fuzzy sets. A fuzzy set is usually defined by its
membership functions. In general, triangular
membership function is used to normalize the crisp
inputs because of its simplicity and computational
efficiency. It is described mathematically in the
manner
12,13
.
Triangle(a,x,y,z)= (1)
Triangle(a,x,y,z)
) 0 ,
y z
a z
,
x y
x a
(min max
|
|
.
|
\
|
=
(2)
where x, y, z are the parameters of the linguistic
value and a is the range of the input parameters.
This triangular membership function as described
in the above expressions (1) and (2) convert the
linguistic values to a range of 0 1. The
membership function of input and output are
represented in Fig. 6 and Fig. 7.
Determination of application rule base
The rules determine input and output membership
functions that will be used in inference process.
These rules are linguistic and also are entitled IF-
THEN rules. In this system, SO
2
have three MF,
NO
2
have three MF,RSPM have three MF, PM
have three MF. The fuzzy system is designed for
predicating Indian AQI (IAQI)
7
and it have four
inputs as SO
2
,NO
2
,RSPM,PM. So the fuzzy
system has 81 rules (3333). Detail of firing
of fuzzy rules is shown in Fig. 8.The proposed
fuzzy model is based on Mamdani fuzzy model
architecture and max-min inference was applied.
The rules of the Mamdani fuzzy system are
generated in Fig. 9.
The fuzzy rules are coded using MATLAB
software as shown in Fig. 9. The Mamdani type
Fuzzy Inference System (FIS) shown in Fig. 5
shows the basic input/output system of the air
quality index model for the rule viewer while a
sample of defuzzified outputs are as shown in
Fig. 10 with surface view in Fig. 11.
Defuzzification of IAQI
After completing the fuzzy decision process, the
fuzzy number obtained must be converted to a
crisp value. This process is entitled defuzzification.
Many method have been developed for
defuzzification. In this study, a centroid (center
of area) technique was applied, which is one of
the most common methods. After defuzzification
process, obtained fuzzy number is geometrical
figure. The crisp value is calculated as
below(equation 3)
14,16
.

}
}

=
dz ) z ( c
dz z ) z ( c
z
(3)
So
2
(3)
No
2
(3)
RSPM (3)
PM (3)
Air Quality
(mamdani)
81 rules
Air Quality (3)
,
x y
x a
,
y z
a z
0,
0,
x a s
y a x s s
z a y s s
a z s
489
Fig. 7 : Membership function of output of FIS
Fig. 6(d) : Membership function of PM
PM
D
e
g
r
e
e

o
f

m
e
m
b
e
r
s
h
i
p
D
e
g
r
e
e

o
f

m
e
m
b
e
r
s
h
i
p
Ai r Quali ty
Fig. 6(b) : Membership function of NO
2
Fig. 6(c) : Membership function of RSPM
D
e
g
r
e
e

o
f

m
e
m
b
e
r
s
h
i
p
RSPM
D
e
g
r
e
e

o
f

m
e
m
b
e
r
s
h
i
p
Fig. 6(a) : Membership function of SO
2
D
e
g
r
e
e

o
f

m
e
m
b
e
r
s
h
i
p
Good
Moderate Poor
1
08
06
04
02
0
0
0 50 100 150 200 250 300 350 400
SO
2
1
08
06
04
02
0
0
0 50 100 150 200 250 300 350 400 500
Good
Moderate Poor
NO
2
1
08
06
04
02
0
0 0 50 100 150 200 250 300 350 400
1
08
06
04
02
0
0
Good
Moderate Poor
Good
Moderate Poor
0 100 200 300 400 500 600 700 800 900
490
Fig. 9 : The rule base coded using MATLAB software.
Fig. 8 : Input output map for the air quality in FIS
Inpur 2
Inpur 3
Inpur 4
491
Fig. 10 : Active rules and membership function input and output of AQI
RESULTS AND DISCUSSION
The prediction of air quality index of five locations
in Chennai,Tamil Nadu,India as per Indian Air
Quality criteria is discussed first. According to
Table 3 to Table 7 represented simulation results.
According to Table 3 to Table 7 , the designed
model predict class A category for air quality
(between 16 to 22 AQI). As per the Table 7, the
designed model predict class B for Vallalar
Nagar(99 AQI) in the month of Augest 2011. From
simulation study it is found that, the air quality of
Adyar(16 AQI) is better compare to Vallalar
Nagar(22 AQI). This indicates that the air quality
is good during the above mentioned period and
the long term exposure may be harmful for the
human beings. From this above simulation result,
it is clear that the proposed FIS model gives very
close results.
Fig. 11 : Surface view of the water quality in FIS
A
i
r

Q
u
a
l
i
t
y
No
2
So
2
492
Table 3 : Simulation result for prediction of AQI in Anna nagar as per Indian
Air Quality Criteria
S/N Month SO
2
NO
2
RSPM PM I-AQ Predication
2011 g/m
3
g/m
3
g/m
3
g/m
3
criteria result of FIS
1. January 8 12 44 48 A 16.5
2. February 10 14 29 51 A 16.25
3. March 8 14 36 40 A 16.27
4. April 7 10 41 40 A 16.41
5. May 7 11 59 58 A 16.98
6. June 7 12 34 50 A 16.22
7. July 8 11 44 46 A 16.5
8. August 7 9 55 32 A 16.84
9. September 8 16 55 32 A 16.84
10. October 7 10 86 28 A 18.05
Table 4 : Simulation result for prediction of AQI in Adyar as per Indian air quality criteria
S/N Month SO
2
NO
2
2011 g/m
3
g/m
3
g/m
3
g/m
3
1. January 7 20 76 112 A 17.62
2. February 10 22 148 127 A 21.16
3. March 10 23 158 187 A 21.75
4. April 9 20 153 136 A 21.46
5. May 11 24 145 144 A 21.01
6. June 9 22 119 177 A 19.62
7. July 9 22 123 149 A 19.82
8. August 8 16 159 111 A 21.79
9. September 8 18 150 130 A 21.29
10. October 7 11 148 97 A 21.16
I-AQ Indian Air Quality, FIS Fuzzy Inference System
493
Table 5 : Simulation result for prediction of AQI in Kilpauk as per Indian Air quality
criteria
S/N Month SO
2
NO
2
2011 mg/m
3
mg/m
3
mg/m
3
mg/m
3
1. January 8 22 127 141 A 20.02
2. February 12 24 110 154 A 19.16
3. March 10 23 102 148 A 18.77
4. April 11 21 119 128 A 19.02
5. May 11 24 155 138 A 21.58
6. June 9 21 107 195 A 19.02
7. July 10 24 92 165 A 18.31
8. August 9 17 141 149 A 20.78
9. September 11 24 193 116 A 21.16
10. October 11 16 189 166 A 22.08
Table 6 : Simulation result for prediction of AQI in T. Nagar as per Indian
air quality criteria
S/N
Month SO
2
NO
2
2011 g/m
3
g/m
3
g/m
3
g/m
3
1. January 7 12 44 155 A 17.69
2. February 12 24 108 128 A 19.07
3. March 9 19 118 127 A 19.56
4. April 9 20 131 112 A 20.25
5. May 10 26 107 144 A 19.02
6. June 9 22 81 132 A 17.83
7. July 9 23 114 114 A 19.36
8. August 8 18 138 92 A 20.61
9. September 11 24 119 110 A 19.62
10. October 9 14 113 85 A 19.31
494
S/N Month SO
2
NO
2
2011 g/m
3
g/m
3
g/m
3
g/m
3
1. January 10 29 134 310 A 21.58
2. February 13 27 151 222 A 21.33
3. March 12 28 178 293 A 22.94
4. April 10 21 121 216 A 19.72
5. May 12 29 133 209 A 20.35
6. June 10 23 137 285 A 20.87
7. July 11 24 196 330 A 22.15
8. August 10 20 272 425 B 99.88
9. September 10 18 203 243 A 19.74
10. October 9 16 166 173 A 22.22
CONCLUSION
In this study, Urban air pollution in many cities is
currently an issue of great concern to the general
public maintaining a high profile on the political
agenda. Anot her reason is t he available
information on the harmful impact of air pollution
on human health, agriculture productivity and
forestry. Despite the concern, little is known about
the mechanisms of chemical transformation and
transportation of air pollutants within the complex
geometry of an urban environment. We have been
assessed air quality of Chennai city with four
pollutants (SO
2
,NO
2
,RSPM,PM) parameters.
This new index is believed to assist decision
makers in reporting the state of air quality,
investigation of spatial and temporal changes. The
authors believe that the fuzzy logic concepts, if
used logically, could be an effective tool for some
of the environmental policy issues. More stringent
methodologies and reliable results are then
required to convince managers and policy makers
to apply fuzzy model in practice
ACKNOWLEDGEMENT
The authors wish to thank Tamil Nadu Pollution
Control Board (TNPCB, Chennai) to provide the
air pollution monitoring data in Chennai, India.
REFERENCES
1. Vallero D. A., Fundamental of air pollution,
academic press, 4th Ed., 574, (2008).
2. Air pollution : What it means for your health.
department of the environment, Transport and
the regions, PO BOX No 236, Wetherby
LS23 7NB,U.K., 156, (1998).
3. Raman Bai V., Bouwmeester and Mohan S.
Fuzzy logic water qualit y Index and
importance of water quality parameters. Air,
Soil. Wat. Res. 2(1), 5159, (2009).
4. Zadeh L. A., Fuzzy Sets. Inform control,
8(1), 338-53, (1965).
5. Ali N., Elshafie A., Othman A. K. and
Othman J., Predication of johon river water
quality parameters using artificial neural
networks. Eur . J. Sci. Res., 28(3), 422-
435,(2009).
6. Mamdani E. H. and Assilian S. , An
experiment in linguistic synthesis with a fuzzy
Table 7 : Simulation of AQI in V. Nagar as per Indian air quality criteria
495
logic controller, Int. J. Man-mach stud.
7(1),1-13, (1975).
7. Takagi T. and Sugeno M. ,Fuzzy identification
of systems and its application to modeling and
Control. IEEE Trans. Syst. Man. cybernet;
15(1), 116-32,(1985).
8. Sugeno M. and Kang G. T. , Structure
identification of fuzzy model, Fuzzy sets. Syst.
28(1),15-33, (1988).
9. Cox E. A. , Fuzzy Fundament als IEEE
Spectrum, 19(10), 58-61(1992).
10. http://home.iitk.ac.in/~mukesh/air-quality/
BASIS.html (2011).
11. http://www.cpcb.nic.in
12. Jang J. S. R., Sun C. T. and Mizutan E.,
Neuro-fuzzy and soft computing. New Delhi;
Prentice, Hall of India Private Limited, 360,
(2005).
13. Mahapatra S.S., Nanda S. K. and Panigrahy
K., A cascaded fuzzy inference system for
Indian river water quality prediction, Advances
in engineering software, Elsevier, 42(1),787-
796, (2011).
14. Nandusekar P. P. , Wagchaure V. M. and
Mahajan Y.S., Measurement of Ambient Air
Quality (AAQ) at ISPAT industries ltd.,
J. Environ. Res. Develop. 3(3), 773-781,
(2009).
15. Ray Soni A.U. and Li Wen W., Health impacts
of traffic related air pollution, J. Environ. Res.
Develop. 4(2), 421-429, (2009).
16. Balashanmugan P., Ramanathan A.R., Elango
E. and Nehrukumar V., Study of ambient air
quality of Chidambaram, Tamilnadu, India, J.
Environ. Res. Develop., 6(2), 365-371,
(2009).
This fight against drilling in the Arctic Refuge is a fight about our principles. Its
about standing up for our environment, our families and our future, and I wont give
up this fight.
Senator John Kerry

10-A Fuzzy Inference System For Air Quality in Using MATLAB, Chennai, India.

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10-A Fuzzy Inference System For Air Quality in Using MATLAB, Chennai, India.

Uploaded by

Copyright:

Available Formats

J. Environ. Res. Develop.

You might also like