FACULTY OF ENGINEERING, D.E.I.

, AGRA
FUZZY SYSTEMS PEE112
Class: M.Tech. Session 2019-20

Syllabus

1
 FACULTY OF ENGINEERING, D.E.I., AGRA
FUZZY SYSTEMS PEE 112
Class: M.Tech. Session: 2019-20

Question Bank
Unit -1
1. State few physical processes where high precision are not required for problem solving.
2. Justify the use of fuzzy logic in support of following statement:
“Requiring precision in engineering models and products translates to requiring high cost and long
lead times in production and development.”
3. Mention chronologically the historical development of Fuzzy Logic.
4. Why uncertainty can be thought of in an epistemological sense as being the inverse of information?
Explain with few exemplary statements, in how many forms uncertainty can be manifested?
5. Give few strong reasons which signify the utility of fuzzy systems theory as problem solving tool
to solve many real life problems.
6. Which of the following statement belongs to fuzzy set or classical (crisp) set? Justify the answer.
Statement A: The set of heights from 5 to 7 feet.
Statement B: The set of heights in the region around 6 feet.
Also represent these sets in the form of mathematical membership functions and membership
diagrams.
7. Develop a reasonable membership function for a square, based on the geometric properties of
rectangle. For this problem use L as the length of the longer side and l as the length of the smaller
side.
8. For the cylindrical shapes shown in Fig.1.1, develop a membership function for each of the
following shapes using the ratio d/h, and discuss the reason for any overlapping among the three
membership functions:
(a) Disc
(b) Cylinder
(c) Rod

(a)
h
(b)

Fig. 1.1

(c)

9. The question of whether a glass of water is half - full or half - empty is an age -old philosophical
issue. Such descriptions of the volume of liquid in a glass depend on the state of mind of the person
asked the question. Develop membership functions for the fuzzy sets "half-full", "empty", and
"half-empty" using percentage volume as the element.
10. Industry A discharges wastewater into a nearby river. Wastewater contains high biological oxygen
demand (BOD) and other inorganic contaminants. The discharge rate of rivers and wastewater is
constant through the year. From research, it has been found that BOD values not exceeding 250
mg/L do not cause any harmful effect to aquatic ecosystems. However, BOD values higher than
250 mg/L have significant impact. Draw both a crisp and fuzzy membership function to show the
effects of the BOD value on aquatic ecosystems.
11. Landfills are a primary source of methane, a greenhouse gas. Landfill caps, called biocaps, are
designed to minimize methane emission by maximizing methane oxidation; these caps are
classified as “best” if they are capable of oxidizing 80% of the methane that originates in the
landfill’s interior. Complete oxidation was found to be difficult to establish. Develop a reasonable
membership function of the percent methane oxidation to show the performance of biocap and
emissions of methane.

2
Unit – 2
1. Define classical (crisp) set. Give few examples of classical sets. How the power set is formed from
the existing universe?
2. (a) What are the operations that can be performed on the classical? Write the expressions involved for
the operations of classical set in function-theoretic form and set-theoretic form.
2. (b) State the properties of classical sets. State Demorgan’s Law. Explain the law with help of Venn
diagram representation.
3. (a) What is the cardinal number of a classical set? How is the cardinality defined for a power set?
3. (b) Discuss in detail how classical sets are mapped to functions.
4. (a) Define fuzzy set. State the reason for membership function to be in the interval 0 to 1.
4. (b) What are the operations that can be performed by a fuzzy set? Explain about the properties present
in fuzzy set.
4. (c) Write the expressions for the fuzzy set operation in set-theoretic form and function theoretic form.
5. (a) How is the excluded middle law different for fuzzy set and the classical set?
5. (b) Discuss about the Demorgen’s law for fuzzy sets. State whether it is similar to that of classical
sets?
6. We want to compare two sensors based upon their detection levels and gain settings. The following
table of gain settings and sensor detection levels with a standard item being monitored provides
typical membership values to represent the detection levels for each of the sensors.
Gain setting Sensor 1 detection level Sensor 2 detection level
0 0 0
20 0.5 0.45
40 0.65 0.6
60 0.85 0.8
80 1 0.95
100 1 1
The universe of discourse is X = { 0, 20, 40, 60, 80, 100} and the membership functions for the
two sensors in standard discrete form are:
𝑆̃1= {0/0 + 0.5/20 + 0.65/40 + 0.85/60 + 1/80 +1/100}
𝑆̃2= {0/0 + 0.45/20 + 0.6/40 + 0.8/60 + 0.95/80 + 1/100}
Find the following membership functions using standard set operations :
(a) 𝑆̃1  𝑆̃2 (b) 𝑆̃1  𝑆̃2 (c) ̅𝑆̃̅̅1̅
(d) ̅𝑆̃̅̅2̅ (d) ̅̅̅̅̅̅̅̅̅̅̅̅
𝑆̃1 ∩ 𝑆̃2 (e) 𝑆̃1  ̅𝑆̃̅̅1̅
̅̅̅̅̅̅̅̅̅̅̅̅
(f) 𝑆̃1 ∪ 𝑆̃2 ̅
(g) 𝑆̃1  𝑆̃1̅̅̅ (h) 𝑆̃2  ̅𝑆̃̅̅2̅
7. For flight simulator data the determination of certain changes in operating conditions of the aircraft
is made on the basis of hard breakpoints in the mach region. Let us define a fuzzy set to represent
the condition of "near" a mach number of 0.74. Further, define a second fuzzy set to represent the
condition of "in the region of" a mach number of 0.74. In typical simulation data a mach number
of 0.74 is hard breakpoint.
𝐴̃ = "near mach 0.74" = { 0/0.734 + 0.8/0.735 + 1/0.74 + 0.6/0.745 + 0/0.75}
𝐵̃ = "in the region of mach 0.74" = {0/0.73 + 0.4/0.735 + 0.8/0.74 + 1/0.745 + 0.6/0.75}
For these fuzzy sets find the following :
(a) 𝐴̃  𝐵̃ (b) 𝐴̃  𝐵̃ (c) 𝐴̃̅
(d) 𝐴̃|𝐵̃ ̅̅̅̅̅̅̅
(e) 𝐴̃ ∪ 𝐵̃ (f) ̅̅̅̅̅̅̅
𝐴̃ ∩ 𝐵̃
8. Suppose an engineer is addressing a problem in the power control of a mobile cellular telephone
̃ be the medium-power fuzzy set, and 𝐻𝑃
transmitting to its base station. Let 𝑀𝑃 ̃ be the high-power
set. Let the universe of discourse be comprised of discrete units of dB.m, i.e., X={0,1,2,……,10}.
The membership functions for these two fuzzy sets are shown in Fig.2.1. For these fuzzy sets,
demonstrate union, intersection, complement and the difference.

3
 µ µ

1 ̃
𝑀𝑃 1 ̃
𝐻𝑃
MP
0 0
x x
0 5 10 0 5 10
Fig. 2.1
9. A system component is tested on a drop table in the time domain t, to shock loads of haversine
pulses of various acceleration amplitudes, d2x/dt2 as shown in Fig. 2.2. After the test the
component is evaluated for damage. Define two fuzzy sets “Passed” = 𝑃̃ and “Failed” = 𝐹̃ . These
sets are define on the real line of |d2x/dt2| which is the magnitude of the input pulse. We define the
following set operations :
(a) 𝐹̃  𝑃̃ = |d2x/dt2| the universe of input shock level results.
(b) 𝐹̃  𝑃̃ : the portion of the universe where the component could both fail and pass.
(c) 𝐹̃̅ : portion of universe that definitely passed.
(d) 𝑃̅̃ : portion of universe that definitely failed.
(e) 𝐹̃ |𝑃̃ : the portion of the failed set that definitely failed.
Define suitable membership functions for the two fuzzy sets 𝐹̃ and 𝑃̃ and determine the
operation just described.
d2x/dt2

Amplitude

෤
𝑃
|d2x/dt2|
t ෤
𝐹

Fig. 2.2
10. (a) Explain about Noninteractive Fuzzy Sets with examples.
(b) Explain following terms with reference to linguistic fuzzy sets:
i) Fuzzy concentration ii) Fuzzy dilation iii) Fuzzy intensification.
11. Three variables of interest in power transistors are the amount of current that can be switched, the
voltage that can be switched, and the cost. The following membership functions for power
transistors were developed from a hypothetical components catalog :
Average current (in amps) = 𝐼̃ = {0.4/0.8 + 0.7/0.9 + 1/1 + 0.8/1.1 + 0.6/1.2}
Average voltage (in volts) = 𝑉̃ = {0.2/30 + 0.8/45 + 1/60 +0.9/75 +0.7/90}
Note how the membership values in each set taper off faster toward the lower voltage and currents.
These two fuzzy sets are related to the "power" of the transistor. Power in electronics is defined by
an algebraic operation, 𝑃̃ = 𝑉̃ 𝐼̃ , but let us deal with a general Cartesian relationship between
voltage and current i.e. simply with 𝑃̃ = 𝑉̃ x 𝐼̃ .
(a) Find the fuzzy Cartesian product 𝑃̃ = 𝑉̃ x 𝐼̃ .
Now let us define a fuzzy set for the cost 𝐶̃ in dollars, of a transistor, e.g.,
𝐶̃ = {0.4/0.5 + 1/0.6 + 0.5/0.7}
(b) Using a fuzzy Cartesian product, find 𝑇̃ =𝐼̃ x 𝐶̃ . What would this relation 𝑇̃, represent
physically.
(c) Using max-min composition, find 𝐸̃ = 𝑃̃ o 𝑇̃ . What would this relation 𝐸̃ , represent physically.
(d) Using max-product composition, find 𝐸̃ = 𝑃̃ o 𝑇̃ .

4
12. The relation between temperature and maximum operating frequency 𝑅̃ depends on various factors
for a given electronic circuit. Let 𝑇̃ be a temperature fuzzy set (in degrees Fahrenheit) and 𝐹̃
represent a fuzzy set (in MHz), on the following universe of discourse :
𝑇̃ = {-100, -50, 0, 50, 100} and 𝐹̃ = {8, 16, 25, 33}
̃ ̃
Suppose a Cartesian product between 𝑇 and 𝐹 is formed that results in the following relation :
-100 -50 0 50 100
8 0.2 0.5 0.7 1 0.9
̃
𝑅= 16 0.3 0.5 0.7 1 0.8
25 0.4 0.6 0.8 0.9 0.4
33 0.9 1 0.8 0.7 0.4
The reliability of the electronic circuit is related TO THE MAXIMUM OPERATING
TEMPERATURE. Such a relation 𝑆̃ can be expressed as Cartesian product between the reliability
index, M = {1, 2, 4, 8, 16} ( in dimensionless units) and the temperature, as in the following
example:
1 2 4 8 16
-100 1 0.8 0.6 0.3 0.1
𝑆̃ = -50 0.7 1 0.7 0.5 0.4
0 0.5 0.6 1 0.8 0.8
50 0.3 0.4 0.6 1 0.9
100 0.9 0.3 0.5 0.7 1
Composition can be performed on any two or more relations with compatible row-column
consistency. To find relationship between frequency and the reliability index, use
(a) max-min composition
(b) max - product composition
13. The formation of algal and other biological colonies in surface waters is strongly dependent on
such factors as the pH of the water, the temperature, and oxygen content. Relationships among
these various factors enable environmental engineers to study issues involving bioremediation
using the algae. Suppose we define a set 𝑇̃ of water temperatures from a lake on the following
discrete universe of temperatures in degrees Fahrenheit as 𝑇̃ = {50, 55, 60}
And Suppose we define a universe of 𝑂̃ of oxygen content values in the water (percent by volume)
as 𝑂̃ = {1, 2, 6}
Suppose a Cartesian product is performed between specific fuzzy sets 𝑇̃ and 𝑂̃ defined on 𝑇̃ and
𝑂̃ to produce the following relation:
1 2 6
𝑅̃ =𝑇̃ x 𝑂̃ = 50 0.1 0.2 0.9
55 0.1 1 0.7
60 0.8 0.7 0.1
Now suppose we define another fuzzy set of temperatures, "about 55 deg. F", with the following
membership values ;
𝐼̃𝑇 = {0.5/50 + 1/55 +0.7/60}
(a) Using max-min composition, find 𝑆̃ = 𝐼̃𝑇 o 𝑅̃ )
(b) Using max-product composition, find 𝑆̃ = 𝐼̃𝑇 o 𝑅̃
14. The accompanying Sagittal diagram show two relations on the universe, {1,2,3}. Are these
relations equivalence relations?

(a) (b)
Fig. 2.3
15. Explain the following in connection with fuzzy set membership function:
(a) Core, support and boundaries.
(b) Normal and subnormal fuzzy sets.

5
 (c) Convex fuzzy set and nonconvex fuzzy set.
(d) Interval valued membership function.
16. (a) What is meant by “Fuzzification” process in fuzzy sets?
(b) Describe Intuition, Inference and Polling concept, methods employed for the membership value
assignment.
(c) Using your own intuition and definitions of the universe of discourse, plot fuzzy membership
functions for:
i) the weight of people. ii) the age of people.
17. Using the inference approach, find the membership values for the triangular shapes isosceles
triangle 𝐼̃, right-angle triangle 𝑅̃ , equilateral triangle 𝐸̃ , isosceles and right-angle triangle 𝐼𝑅
̃ and
̃
other triangles 𝑇 with a triangle with angles 80°, 55° and 45°.
18. (a) How is the polling concept adopted in rank ordering method to define the membership values?
(b) Suppose 1000 people respond to a questionnaire about their pairwise preferences among five
cars, X = {Maruti 800, Scarpio, Matiz, Santro, Octavia} as shown in following table. The summary
of opinion survey shows that out of 1000 people, 192 preferred Maruti 800 to the Scarpio, etc.
Define a fuzzy set 𝐴̃ on the universe of cars “best car” using Rank Ordering method.
Number who preferred
Maruti 800 Scarpio Matiz Santro Octavia Total
Maruti 800 - 192 246 592 621 1651
Scarpio 403 - 621 540 391 1955
Matiz 235 336 - 797 492 1860
Santro 523 364 417 - 608 1912
Octavia 616 534 746 726 - 2622
Total 10000

Unit-3
1. Explain about Hartley information and Shennon entropy, classical measures of uncertainty.
2 Let the set X = {x1, x2, x3, x4, x5} with the probability distribution:
p = (p1 = 0.25, p2 = 0.5, p3 = 0.0625, p4 = 0.0625, p5 = 0.125) is given then calculate the
uncertainty in terms of the Shannon entropy.
3. Under what conditions of P and Q is the implication P →Q a tautology?
4. Explain modus ponens inference and modus tollens inference used in propositional logic and
construct their respective truth table.
5. The exclusive-or is given by the expression, P XOR Q = (P ̅ ᴧ Q)ᴠ (PᴧQ
̅ ) . Show that the logical-
or, given by P ᴠ Q, gives a different results from the exclusive-or and comment on this difference
using an example in your own field.
6. For a proposition R of the form P→ Q, show the following :
(a) R and its contrapositive are equivalent, i.e. prove that (P→ Q)  (Q ̅→ ̅ P ).
(b) The converse of R and the inverse of R are equivalent, i.e. prove that (Q→P)  (P ̅ →Q ̅ ).
7. ̅ ̅
Show that the dual of the equivalence (P ᴠ Q)ᴠ (P ᴧQ)  X is also true.
8. Prove that the following statements by contradiction.
(a) (( P → Q) ᴧ P) → Q
̅ ) ᴧ (Q ᴠ R
(b) ((P → Q ̅ ) ᴧ (R ᴧ S̅ )) → ̅P
9. Two universes of discourse for a heat exchanger problem are described by the following
collection of elements, X={1,2,3,4} and Y={1,2,3,4,5,6}. Suppose X is a universe of normalized
temperatures and Y is a universe of normalized pressures.
Crisp set A on universe X and crisp set B on universe Y are defined as A={2,3}or A=
0 1 1 0 0 0 1 1 0 0
{1 + 2 + 3 + 4} (using Zadeh’s notation) and B={3,4} or B= {1 + 2 + 3 + 4 + 5 + 6} (using
Zadeh’s notation) respectively.
(i) Find the matrix R that represents the rule IF A, THEN B as matrix of characteristic (crisp
membership) values.
(iii) Suppose a crisp set C on the universe of normalized temperatures Y is defined as C = {5,6}.
Find the full relation R describing the implication IF A, THEN B, ELSE C.
10. Suppose commercial potential of a new invention is to be evaluated using following criterion:
The two metrics are used to make decisions regarding the innovation of the idea. The metrics
are the “uniqueness” of the invention, denoted by a universe of novelty scales, X={1,2,3,4}, and
the “market size” of the invitation’s commercial market, denoted on a universe of scaled market

6
 sizes, Y={1,2,3,4,5,6}. In both universes the lowest numbers are the “highest uniqueness” and
the “largest market,” respectively.
A new invention, say a compressible liquid of very useful temperature and viscosity conditions,
has just received scores of “medium uniqueness,” denoted by fuzzy set A ̃ , and “medium market
size,” denoted by fuzzy set B̃. An invention is assigned the following fuzzy sets to represent its
ratings:
̃ = medium uniqueness = {0.6 + 1 + 0.2}
A 2 3 4
̃ = medium market size = {0.4 + 1 + 0.8 + 0.3}
B 2 3 4 5
̃ THEN B
Determine the implication of such a result, i.e., IF A ̃.
11. A soils engineer wishes to track the movement of soil particles under applied loading in an
experimental apparatus that allows viewing of the soil motion. Suppose soil engineer is building
pattern recognition software to enable a computer to monitor and detect the motions. However,
there are some difficulties in “teaching” the software. The occlusion can occur when a tracked
particle is behind another particle, behind a mark on the camera’s lens, or partially out of sight
of the camera. We want to establish a relationship between particle occlusion, which is a poorly
known phenomenon, and lens occlusion, which is quite well-known in photography. Let these
membership functions,
̃ = {0.1 + 0.9 + 0.0} and B
A ̃ = { 0 + 1 + 0 } describe fuzzy sets for a tracked particle moderately
𝑥
1 𝑥
2 𝑥3 𝑦 1𝑦 𝑦 2 3
occluded behind another particle and a lens mark associated with moderate image quality,
̃ is defined on a universe X = {x1, x2, x3} of tracked particle indicators,
respectively. Fuzzy set A
and fuzzy set B ̃ (which is crisp singleton) is defined on a universe Y = {y1, y2, y3} of lens
obstruction indices. Determine the implication of following rule:
IF occlusion due to a particle is moderate, THEN image quality will be similar to a moderate lens
obstruction, or symbolically,
̃ , THEN y is B
IF x is A ̃.
12. Consider the following two discrete fuzzy sets, which are defined on universe X={-5,5}:
̃ = “zero” = { 0 + 0.5 + 1.0 + 0.5 + 0}
A −2 −1 0 1 2
̃ = “positive medium” = {0 + 0.5 + 1.0 + 0.5 + 0}
B 0 1 2 3 4
(a) ̃ , THEN B
Construct the relation for the rule IF A ̃ (i.e., IF x is “zero” THEN y is “positive
medium”) using the Mamdani implication, 𝜇𝑅̃ (x,y)= min[𝜇𝐴̃(x), 𝜇𝐵̃ (y)], and the product
implication 𝜇𝑅̃ (x,y) = 𝜇𝐴̃(x).𝜇𝐵̃ (y)
(b) If a new antecedent is introduced as,
̃ ’ = “positive small” = { 0 + 0.5 + 1.0 + 0.5 + 0}
A −1 0 1 2 3
Find the new consequent B′ ̃ , using max-min composition, i.e., B′ ̃ = A ̃ ′ ͦ 𝑅̃ ,
for both relations from part (a).
13. Explain the fundamental terms (atomic terms) and composite terms with examples used in our
natural language. How linguistic hedges (modifiers) are used in fuzzy sets.
14. Suppose we have a universe of integers, Y = {1,2,3,4,5}. We define the following linguistic terms
as mapping onto Y:
1 0.8 0.6 0.4 0.2
“Small” = {1 + 2 + 3 + 4 + 5 }
0.2 0.4 0.6 0.8 1
“Large” = { 1 + 2 + 3 + 4 + 5}
i) Modify these two linguistic terms with hedges “Very small”, “Not very small”
ii) Construct a phrase, or a composite term, “not very small and not very, very large”.
15. (a) Explain following techniques used for decomposition of linguistic rules involving multiple
antecedents into the simple canonical form
i) Multiple conjunctive antecedents
ii) Multiple disjunctive antecedents
(b) Discuss the methods of aggregation of fuzzy rules.
16. Describe in detail of formulation of inference rules in Mamdani Fuzzy Inference System (FIS).
17. The fuzzy sets A ̃, B
̃, and C̃ are all defined on the universe X =[ 0,5] with the following
membership functions :
𝜇𝐴̃ (x) = 1/(1+5(x-5)2) 𝜇𝐵̃ (x) = 2-x 𝜇𝐶̃ (x) = 2x/(x+5)
(a) Sketch the membership functions.

7
 (b) Define the intervals along the x axis corresponding to the -cut sets for each of the fuzzy
̃, B
sets A ̃, and C̃ for the following values of  :
(i)  =0.2
(ii)  =0.4
(iii) =0.7
(iv)  =0.9
(v)  = 1.0
18. Explain the following methods to perform defuzzification process:
i. Max membership principle or height method.
ii. Centroid method or center of area method or center of gravity method.
iii. Weighted average method.
iv. Mean max membership.
19. Find the defuzzified value by weighted average method shown in Fig. 3.1.

Fig. 3.1
20. By using centroid method of defuzzifiation convert fuzzy value x to precise value Z* for the
following Fig. 3.2.

Fig. 3.2
21. In heat exchanger design, a flexibility analysis requires the designer to determine if the size of
the heat exchanger is either small or large. In order to quantify this linguistic vagueness of size,
the general design equation for a heat exchanger is formed as Q = AU ΔTlog mean, where the heat
transfer coefficient U and area A need to be determined. Fig. 3.3 shows a schematic of this
exchanger.

Fig. 3.3 Heat exchanger design

Suppose we want to determine the sizes of a heat exchanger in which a steam of benzene is heated
using saturated steam at pressure 68.95 kPa and temperature 362.7K. The initial temperature of
the benzene steam is 17◦ C, and the model used to determine the size of the heat exchanger is the
following:
𝑇𝑆 − 𝑇1
𝐴𝑈 = 𝑤𝐶𝑝 𝑙𝑛 ( )
∆𝑇𝑎𝑝𝑝
Where Cp is the heat capacity of the benzene (1.7543 Kj/K kg) and TS-T1 = 72.55 K.
Benzene flow rate, w, is considered in kg/s, and temperature approach (ΔTapp) in kelvin, are
modelled as the inputs, and the size of the heat exchanger (AU) is modelled as output. The
following disjunctive rules of inference based on observations of the model are as follows:

8
 Rule 1: If w is 𝐴̃11 (large flow rate) and ΔTapp is 𝐴̃12 (small approach), THEN AU is B
̃1 (large heat
exchanger).
Rule 2: If w is 𝐴̃21 (small flow rate) or ΔTapp is 𝐴̃22 (large approach), THEN AU is B
̃1 (small heat
exchanger).
Rule 3: If w is 𝐴̃21 (small flow rate) and ΔTapp is 𝐴̃12 (small approach), THEN AU is B
̃1 (large heat
exchanger).

Fig. 3.4 Graphical inference using the Mamdani method for three rules
The graphical equivalent of these rules is shown in Fig. 3.4. Obtain heat exchanger size (AU*) if
a weighted average defuzzification method is employed for the following input pair using
Mamdani inference method.
W = 1300 kg/s and ΔTapp = 6.5 K

Unit-IV
1. Explain mathematical model for general problem of feedback control system design. Write
different mathematical forms for
i) Proportional (P) controller.
ii) Proportional-plus-integral (PI) controller.
iii) Proportional-plus-derivative (PD) controller.
iv) Proportional-plus-derivative-plus-integral (PID) controller.
2. State the assumptions made in Fuzzy Control System Design?
3. Explain the steps in designing a simple fuzzy control system with help of a block diagram of
simple fuzzy logic control system.
4. Explain about choice of process state, control output variables, content of the rule antecedent and
rule consequent for each of the rule in designing P, PD, PI and PID like fuzzy knowledge based
controller.
5. (a) Explain how is fuzzy controller different to conventional controller?
(b) Is fuzzy controller nonlinear ? If yes, explain the reasons.
6. Explain qualitatively the influence of membership function and cross points on fuzzy
controllers.
7. Explain the influence of symmetry and width of membership function and noise in construction
of data base of fuzzy controller.
8. Discuss the role of choice of scaling factor for fuzzy knowledge based controller.
9. Explain the influence of choosing type of defuzzification method on fuzzy controller.
10. The system as illustrated in Fig. 4.1 comprises a dial to control the flow of warm/hot or
cool/cold air and thermometer to measure the room temperature (To C). When the dial is turned
positive, warm/hot air is supplied from the air conditioner and if it is turned negative, cool/cold
air is supplied. If set to zero, no air is supplied.

9
 Fig. 4.1 Air conditioner control system
A person now notices the difference in temperature (ΔTo C) between the room temperature (To
C) as measured by the thermometer and the desired temperature (TOo C) at which the room is
desired to be kept (set-point).
Determine to what extent the dial should be turned so that the appropriate supply of air
(hot/warm/cool/cold) will nullify the change in temperature.
The rule base for above problem is given in the following table.

S.No. Fuzzy rule (Descriptive) Fuzzy rule (Notational)

1. If the room temperature is approximately equal to the set point 𝑑∆𝑇
If ∆𝑇 is ZE and 𝑑𝑡 is PL then dial
To C, ΔT is approximately Zero (ZE)
𝑑∆𝑇 should be NL.
and the temperature is rapidly changing higher, i.e. 𝑑𝑡 is
positively large (PL) then blow cold air rapidly, i.e. turn the dial
negative large (NL).
2. If the room temperature is high 𝑑∆𝑇
If ∆𝑇 is PL and is ZE then dial
𝑑𝑡
and there is no change in temperature, i.e. ΔT is positive large
𝑑∆𝑇 should be NM.
(PL) and. 𝑑𝑡 is approximately zero (ZE)
then blow cold air at intermediate level, i.e. turn the dial
negative medium (NM).
3. If the room temperature is a little bit higher than the set point 𝑑∆𝑇
If ∆𝑇 is PS and is NS then dial
𝑑𝑡
and the temperature is gradually decreasing, i.e. ΔT is positively
𝑑∆𝑇 should be ZE.
small (PS) and 𝑑𝑡 is negatively small (NS)
then there is no need to blow hot or cold air, i.e. turn the dial to
approximately zero (ZE).
𝑑∆𝑇
The fuzzy sets for the system inputs, namely ∆𝑇 and and the system output, namely turn of
𝑑𝑡
the dial are as shown in Fig. 4.2.
μ
μ μ
1 PL
1 ZE 1 PS

∆𝑇:

-5 0 5 ∆𝑇
-5 0 5 ∆𝑇 -5 0 5 ∆𝑇

μ μ μ
1 PL 1 ZE 1 NS
𝑑∆𝑇
:
𝑑𝑡

-3°C/min 3°C/min -3°C/min 3°C/min -3°C/min 3°C/min

𝑑∆𝑇 𝑑∆𝑇 𝑑∆𝑇
𝑑𝑡 𝑑𝑡 𝑑𝑡

μ μ μ
NL 1 NM 1 1 ZE

Dial turn
z:

-1 0 +1 -1 0 +1 -1 0 +1
z z z

Fig. 4.2 Fuzzy sets for the air conditioner control system
𝐝∆𝐓
Consider the system inputs, ΔT = 2.5°C and 𝐝𝐭 = -1°C/min and defuzzification using weighted
average method.
10
Unit-V
1. Explain about Multi-objective Decision Making using fuzzy systems approach.
2. A geotechnical engineer on a construction project must prevent a large mass of soil from sliding
into a building site during construction and must retain this mass of soil indefinitely after
construction to maintain stability of the area around a new facility to be constructed on the site.
The engineer therefore must decide which type of retaining wall design to select for the project.
Among the many alternative designs available, the engineer reduces the list of candidate retaining
wall designs to three: (1) a mechanically stabilized embankment (MSE) wall, (2) a mass concrete
spread wall (Conc), and (3) a gabion (Gab) wall. The owner of the facility (the decision maker)
has defined four objectives that impact the decision: (1) the cost of the wall (Cost), (2) the
maintainability (Main) of the wall, (3) whether the design is a standard one (SD), and (4) the
environmental (Env) impact of the wall. Moreover, the owner also decides to rank the preferences
for these objectives on the unit interval. Hence, the engineer sets up the problem as follows:
A = {MSE, Conc, Gab} = {a1, a2, a3}
O = {Cost, Main, SD, Env} = {O1,O2,O3,O4}
P = {b1, b2, b3, b4} −→[0, 1]
From previous experience with various wall designs, the engineer first rates the retaining walls
with respect to the objectives, given here. These ratings are fuzzy sets expressed in Zadeh’s
notation.

O1 =  0.4
+
1
+
0.1

MSE Conc Gab
O2 =  0.7
+
0.8
+
0.4

MSE Conc Gab
O3 =  0.2
+
0.4
+
1

MSE Conc Gab
O2 =  1
+
0.5
+
0.5

MSE Conc Gab
The preferences are given as follows:
b1 = 0.8 b2 = 0.9 b3 = 0.7 and b4 = 0.5
3. An aluminium company needs to improve the performance of its cold rolling mill in order to
meet new customer specifications. One way to improve the mill performance is to keep the mill
equipment and to keep the motor drives, cylinder and control systems that control the final
thickness of the metal of the mill. The alternatives being considered for upgrading the control
system are (i) developing and installing an in-house proprietary system (IN); (ii) developing a
specification and getting an outside vendor to develop and install the system (OUT); and (iii) just
keep the current system and make the minor modifications (OLD).
The company has several objectives that it would like to meet with respect to the project. The
first is the COST-it wants to minimize this. The second is the FLEXIBILITY-it wants the ability
to make the future changes and respond quickly to meet the market demands. The third is the
DOWNTIME-the company wants to minimize the downtime associated with upgrading to reduce
loss production time. The fourth is the PERFORMANCE against new market tolerances-it needs
to meet new market specifications. The project team which includes a control engineer,
production manager, and mechanical engineer, rate the alternatives against the objectives as:

COST ̃1 =
𝑂  0 .4 0.6
+ +
1

IN OUT OLD
̃2 =  
0 .9 0.7 0 .2
FLEXIBILITY 𝑂 + +
IN OUT OLD
̃3 =  
0 .5 0.5 1
DOWNTIME 𝑂 + +
IN OUT OLD
̃4 =  
1 0.8 0 .2
PERFORMANCE 𝑂 + +
IN OUT OLD
Preferences define by the company are COST, 0.6; FLEXIBILITY,0.7; DOWN-TIME,0.8; and
PERFORMANCE, 0.9.Determine the best alternative for upgrading the control system.
11
4. A person is responsible for selecting an electronic CAE package to do schematic capture, printed
circuit board design, and design simulation. The preliminary selection comes up with the high
performance packages Newton Graphics (NEWT), cadence (CAD), and Racal Redac (RAC). The
objective that effect the final decision are (i) initial cost of the packages, (ii) maintenance costs,
(iii) hardware and operating system support, and (iv) speed of processing a milestone model.
Based on the prior experience on maintenance costs and performance, the objectives are ranked
as follows:
̃
𝑂𝑐𝑜𝑠𝑡 =  0.4
+
1
+
0 .6

NEWT CAD RAC
𝑚𝑎𝑖𝑛𝑡 =  
0.5 0.7 0 .6
𝑂̃ + +
NEWT CAD RAC
̃ =  
1 0.6 0 .8
𝑂ℎ𝑎𝑟𝑑𝑤𝑎𝑟𝑒 + +
NEWT CAD RAC
𝑠𝑝𝑒𝑒𝑑 =  
0.9 0.7 0 .6
𝑂̃ + +
NEWT CAD RAC
The preferences assigned to each objectives are- b1 = 0.9,b2 = 0.7,b3=0.5,b4 = 0.5. Determine the
CAE package that is suitable for the project.
5. Evaluate three different approaches to controlling conditions of an aluminum smelting cell (with
respect to voltage across the cell and alumina concentration in the path). The control approaches
are-
a1 = AGT: aggressive control tuning (very reactive)
a2 = MOD: moderate control tuning (mildly reactive)
a3 = MAN: essentially manual operation (very little computer control)
There are several objectives to consider:
̃1 : minimum power consumption
𝑂
̃2 : overall operating stability
𝑂
̃3 : minimum environmental impact
𝑂
The control approaches are related as follows:
 0.7 0.6 0.3 
̃1 =
𝑂  + + 
 AGT MOD MAN 
 0.45 0.8 0.6 
̃2 =
𝑂  + + 
 AGT MOD MAN 
 0.5 0.62 0.4 
̃3 =
𝑂  + + 
 AGT MOD MAN 
The preferences are given by- b1 = 0.8, b2 = 0.5, b3 = 0.6
What is the best choice of the control?
6. One of the most important activities of a ship owner is the preservation of its ships. One of the
factors in preserving a ship in a serviceable condition is painting since sea water is highly
corrosive. The criteria used by ship owners in selecting marine paints are (i) corrosion resistance
(COR), (ii) durability (DUR), (iii) availability (AV), (iv) toxicity (TOX), and (v) cost. A shipping
company is presented with five choices of paint (B i, i = 1, 5). The company needs to select one
of the brands so that the purchasing department can negotiate a price with the paint
manufacturers. The engineering department evaluated the five brands of paint with respect to the
objectives as follows.
 0.4 0.7 0.8 0.6 0.3 
̃
𝑂𝐶𝑂𝑅 =  + + + + 
 B1 B2 B3 B4 B5 
 0.5 0.9 0.5 0.8 0.2 
̃
𝑂𝐷𝑈𝑅 =  + + + + 
 B1 B2 B3 B4 B5 
 0.9 0.7 0.4 0.3 0.7 
𝑂̃
𝐴𝑉 =  + + + + 
 B1 B2 B3 B4 B5 
12
  0.6 0.9 0.8 0.3 0.9 
̃
𝑂𝑇𝑂𝑋 =  + + + + 
 B1 B2 B3 B4 B5 
 0.3 0.4 0.2 0.5 0.4 
𝑂̃
𝐶𝑂𝑆𝑇 =  + + + + 
 B1 B2 B3 B4 B5 
The ship owner ranked the purchasing objectives as corrosion = 0.9, durability = 1.0, availability
= 0.6, toxicity = 0.7, and cost = 0.3. Find which brand of the paint the shipping company should
select for its ship preservation program.
7. A carcinogen, trichloroethylene (TCE), has been detected in soil and ground water at levels
higher than the EPA maximum contaminant levels (MCLs). There is an immediate need to
remediate soil and ground water. Three remediation alternatives-(i) pump and treat with air
stripping (PTA), (ii) pump and treat with photo-oxidation (PTP), and (iii) bioremediation of soil
with pump and treat and air stripping (BPTA)-are investigated.
The objectives are these: cost (O1), effectiveness ((O2, capacity to reduce the contaminant
concentration), duration (O3), and speed of implementation (O4). The ranking of the alternatives
on each objective are given as follows:
 0.7 0.9 0.3 
̃1 =
𝑂  + + 
 PTA PTP BPTA
 0.4 0.6 0.8 
̃2 = 
𝑂 + + 
 PTA PTP BPTA
 0.7 0.3 0.6 
̃3 =
𝑂  + + 
 PTA PTP BPTA
 0.8 0.5 0.5 
̃4 = 
𝑂 + + 
 PTA PTP BPTA
The preferences for each objective are P = 0.6,0.8,0.7,0.5. Determine the optimum choice
of a remediation alternative.
8. What is system identification problem? Explain a method reported by Tsukamato and Terano
that produces interval values for the solution.
9. A fuzzy relation has an expression given as
0.4 0.6
{a1 , a2} o [ ] = [0.3 0.1]
0.8 0.1
Find ai ( i = 1,2 ) by using Tsukamato’s method for inverse fuzzy relations.
10. Consider a suitable example and explain about nonlinear simulation using fuzzy systems using
fuzzy relational equations.

********

13