• Home
  • Textbooks
  • Fuzzy Logic with Engineering Applications
  • Classical Relations and Fuzzy Relations

Fuzzy Logic with Engineering Applications

Timothy Ross

Chapter 3

Classical Relations and Fuzzy Relations - all with Video Answers

Educators


Chapter Questions

Problem 1

The provision of high-quality drinking water remains one of the greatest environmental challenges for public health officials and water utilities worldwide. In order to ensure maximum water quality in distribution networks, proper attention must be given to the pressure head at nodal points (measured by pressure probes) and the demand consumption pattern (measured by telemetry meters) along the whole length of the distribution network. Suppose we have two fuzzy sets, P defined on a universe of three discrete pressures $\left\{x_1, x_2, x_3\right\}$, and $\underset{D}{ }$ defined on a universe of two discrete demand consumptions $\left\{y_1, y_2\right\}$, where fuzzy set $\stackrel{P}{\sim}$ represents the near optimum pressure for high-quality drinking water and $P$ represents the instantaneous demand (water demand) obtained from a time-series demand forecasting. Thus, $\underset{\sim}{\mathrm{P}}$ and $\underset{\sim}{\mathrm{D}}$ represent the variable inputs and water quality represents the output.

The Cartesian product represents the conditions (pressure-demand consumption) of the distribution system that are associated with near maximum water quality.

Let

$$
\stackrel{\mathrm{P}}{\sim}=\left\{\frac{0.1}{x_1}+\frac{0.4}{x_2}+\frac{1}{x_3}\right\} \quad \text { and } \quad \underset{\sim}{\mathrm{D}}=\left\{\frac{0.5}{y_1}+\frac{0.8}{y_2}\right\}
$$

Check back soon!

Problem 2

In a water treatment process, we use a biological process to remove biodegradable organic matter. The organic matter is measured as the biological oxygen demand (BOD), where the optimal BOD of effluent should be less them $20 \mathrm{mg} / \mathrm{L}$. Let B represent a fuzzy set "good effluent" on the universe of optical BOD values $(20,40,60)$ as defined by the membership function

$$
\mu_{\underline{B}}=\frac{0.5}{60}+\frac{0.7}{40}+\frac{1.0}{20}
$$

The retention time is critical to a bioreactor; we try to find the retention time, measured in days. Let T represent a fuzzy set called "optimal retention time" on the universe of days $(6,8,10)$ as given by the membership function

$$
\mu_{\mathrm{T}}=\frac{0.9}{10}+\frac{0.7}{8}+\frac{0.5}{6}
$$

The utilization rate of organic food indicates the level of the treatment in the biological process, and this rate is measured on a universe of fractions from 0 to 1 , where 1 is optimal. Fuzzy set $\underset{\sim}{\mathrm{U}}$ will represent "high utilization rates," as defined by the membership function

$$
\mu_{\mathrm{U}}=\frac{1}{0.9}+\frac{0.8}{0.8}+\frac{0.6}{0.7}+\frac{0.4}{0.6}
$$

We can define the following relations:
$\mathrm{R}=\mathrm{B} \times \mathrm{T}$, which reflects how retention time affects BOD removal;
$\mathrm{S}=\mathrm{T} \times \mathrm{U}$, which relates how retention time affects organic food consumption; and
$W=R \circ S$, which represents the BOD removal and food utilization.
(a) Find R and $\underset{\sim}{\mathrm{S}}$ using Cartesian products.
(b) Find $W$ using max-min composition.
(c) Find $\underset{W}{W}$ using max-product composition.

Check back soon!

Problem 3

Assume storm magnitudes are recorded on a rain gauge station within a 24 h period. We will represent our assessment of the size of a storm on the universe of rainfall depths, $h_i$, $i=1,2,3$, where $h_3>h_2>h_1$. The data on depths are based on statistical estimates acquired from numerous rainfall records. The membership function representing the confidence in the rainfall depth of a particular "moderate storm" F is given by

$$
\underset{F}{F}=\left\{\frac{0.4}{h_1}+\frac{0.9}{h_2}+\frac{0.6}{h_3}\right\}
$$

Suppose D is a fuzzy set which represents the rainfall duration, $t_i\left(t_i<24 \mathrm{~h}\right)$, where $t_2>t_1$ and the duration can again be derived from statistics. The membership function of a "long duration storm" might be

$$
\mathrm{D}=\left\{\frac{0.1}{t_1}+\frac{1.0}{t_2}\right\}
$$

(a) Find the Cartesian product $\underset{\sim}{\mathrm{F}} \times \underset{\sim}{\mathrm{D}}=\underline{\mathrm{G}}$, which provides a relation between rainfall depth and duration.
(b) Then assume you have a fuzzy set of confidence in the measurement of the rainfall depth due to factors such as wind, human error, instrument type, etc. Such a fuzzy set on the universe of depths, say "high confidence in depth $h_2{ }^{\prime \prime}$, could be

$$
\mathrm{E}=\left\{\frac{0.2}{h_1}+\frac{1.0}{h_2}+\frac{0.3}{h_3}\right\}
$$

Using a max-min composition find $\underset{\sim}{\mathrm{C}}=\underset{\mathrm{E}}{\mathrm{O}} \mathrm{G}$, which represents the best strength of the estimate with respect to the storm duration.

Check back soon!

Problem 4

A company sells a product called a video multiplexer, which multiplexes the video from 16 video cameras into a single video cassette recorder (VCR). The product has a motion detection feature that can increase the frequency with which a given camera's video is recorded to tape depending on the amount of motion that is present. It does this by recording more information from that camera at the expense of the amount of video that is recorded from the other 15 cameras. Define a universe X to be the speed of the objects that are present in the video of camera 1 (there are 16 cameras). For example, let $\mathrm{X}=$ \{Low Speed, Medium Speed, High Speed $\}=$ \{LS, MS, HS\}. Now, define a universe Y to represent the frequency with which the video from camera 1 is recorded to a VCR tape, i.e., the record rate of camera 1. Suppose $\mathrm{Y}=$ \{Slow Record Rate, Medium Record Rate, Fast Record Rate $\}=\{$ SRR, MRR, FRR $\}$. Let us now define a fuzzy set $A$ on $X$ and a fuzzy set $B$ on $Y$, where $A$ represents a fuzzy slow-moving object present in video camera 1 , and $\underset{\sim}{B}$ represents a fuzzy slow record rate, biased to the slow side. For example,

$$
\begin{aligned}
& \mathrm{A}=\left\{\frac{1}{\mathrm{LS}}+\frac{0.4}{\mathrm{MS}}+\frac{0.2}{\mathrm{HS}}\right\} \\
& \mathrm{B}=\left\{\frac{1}{\mathrm{SRR}}+\frac{0.5}{\mathrm{MRR}}+\frac{0.25}{\mathrm{FRR}}\right\}
\end{aligned}
$$

(a) Find the fuzzy relation for the Cartesian product of $\underset{\sim}{A}$ and $B$, i.c., find $R=A \times B$.
(b) Suppose we introduce another fuzzy set, C, which represents a fuzzy fast-moving object present in video camera 1 , say, for example, the following:

$$
\mathrm{C}=\left\{\frac{0.1}{\mathrm{LS}}+\frac{0.3}{\mathrm{MS}}+\frac{1}{\mathrm{HS}}\right\}
$$

Find the relation between $\underset{\sim}{\mathrm{C}}$ and $\underset{\sim}{\mathrm{B}}$ using a Cartesian product, i.e., find $\underset{\sim}{\mathrm{S}}=\underset{\sim}{\mathrm{C}} \times \underset{\sim}{\mathrm{B}}$.
(c) Find $\mathrm{C}^{\circ} \mathrm{R}$ using max-min composition.
(d) Find CoR using max-product composition.
(e) Comment on the differences between the results of parts (c) and (d).

Check back soon!

Problem 5

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:

$$
\begin{aligned}
& \text { Average current (in amps) }=\underset{\sim}{I}=\left\{\frac{0.4}{0.8}+\frac{0.7}{0.9}+\frac{1}{1}+\frac{0.8}{1.1}+\frac{0.6}{1.2}\right\} \\
& \text { Average voltage (in volts) }=\underset{V}{V}=\left\{\frac{0.2}{30}+\frac{0.8}{45}+\frac{1}{60}+\frac{0.9}{75}+\frac{0.7}{90}\right\}
\end{aligned}
$$

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, $P=V I$, but let us deal with a general Cartesian relationship between voltage and current, i.e., simply with $\underline{\mathrm{P}}=\underline{V} \times 1$. Keep in mind that the Cartesian product is different from the arithmetic product. The Cartesian product expresses the relationship between $V_i$ and $I_j$, where $V_i$ and $I_j$ are individual elements in the fuzzy sets V and I .
(a) Find the fuzzy Cartesian product $\mathrm{P}=\mathrm{V} \times \underline{I}$.

Now let us define a fuzzy set for the cost C , in dollars, of a transistor, e.g.,

$$
\mathrm{C}=\left\{\frac{0.4}{0.5}+\frac{1}{0.6}+\frac{0.5}{0.7}\right\}
$$

(b) Using a fuzzy Cartesian product, find $\mathrm{T}=\mathrm{I} \times \underset{\sim}{\mathrm{C}}$. What would this relation, T , represent physically?
(c) Using max $-\min$ composition, find $\underset{\sim}{\mathrm{E}}=\underset{\sim}{\mathrm{P}} \mathrm{\sim} \mathrm{T}$. What would this relation, E , represent physically?
(d) Using max-product composition, find $\mathrm{E}=\mathrm{P} \circ \mathrm{T}$.

Check back soon!

Problem 6

The relationship between temperature and maximum operating frequency R depends on various factors for a given electronic circuit. Let $T$ be a temperature fuzzy set (in degrees Fahrenheit) and F represent a frequency fuzzy set (in $\overline{\mathrm{MHz}}$ ) on the following universes of discourse:

$$
\mathrm{T}=\{-100,-50,0,50,100\} \text { and } \mathrm{F}=\{8,16,25,33\}
$$

Suppose a Cartesian product between T and $\underset{\sim}{\mathrm{F}}$ is formed that results in the following relation:

$$
\underset{R}{R} \begin{array}{r}
8 \\
16 \\
25 \\
33
\end{array}\left[\begin{array}{ccccc}
-100 & -50 & 0 & 50 & 100 \\
0.2 & 0.5 & 0.7 & 1 & 0.9 \\
0.3 & 0.5 & 0.7 & 1 & 0.8 \\
0.4 & 0.6 & 0.8 & 0.9 & 0.4 \\
0.9 & 1 & 0.8 & 0.6 & 0.4
\end{array}\right]
$$
The reliability of the electronic circuit is related to the maximum operating temperature. Such a relation S can be expressed as a Cartesian product between the reliability index, $\underline{M}=\{1,2,4,8,16\}$ (in dimensionless units), and the temperature, as in the following example:

$$
\mathrm{S}=\begin{array}{r}
-100 \\
-50 \\
0 \\
50 \\
100
\end{array}\left[\begin{array}{ccccc}
1 & 2 & 4 & 8 & 16 \\
1 & 0.8 & 0.6 & 0.3 & 0.1 \\
0.7 & 1 & 0.7 & 0.5 & 0.4 \\
0.5 & 0.6 & 1 & 0.8 & 0.8 \\
0.3 & 0.4 & 0.6 & 1 & 0.9 \\
0.9 & 0.3 & 0.5 & 0.7 & 1
\end{array}\right]
$$

Composition can be performed on any two or more relations with compatible row-column consistency. To find a relationship between frequency and the reliability index, use
(a) max-min composition
(b) max-product composition

Check back soon!

Problem 7

The formation of algal solutions 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 T of water temperatures from a lake on the following discrete universe of temperatures in degrees Fahrenheit:

$$
\mathrm{T}=\{50,55,60\}
$$

And suppose we define a universe $O$ of oxygen content values in the water, as percent by volume:

$$
O=\{1,2,6\}
$$

Suppose a Cartesian product is performed between specific fuzzy sets T and $\underset{\sim}{O}$ defined on T and Q to produce the following relation:

$$
\left.R=\mathrm{T} \times \mathrm{O}=55 \begin{array}{c}
50 \\
60
\end{array} \begin{array}{ccc}
1 & 2 & 6 \\
0.1 & 0.2 & 0.9 \\
0.1 & 1 & 0.7 \\
0.8 & 0.7 & 0.1
\end{array}\right]
$$

Now suppose we define another fuzzy set of temperatures, "'about $55^{\circ} \mathrm{F}$, '' with the following membership values:

$$
\mathrm{I} r=\left\{\frac{0.5}{50}+\frac{1}{55}+\frac{0.7}{60}\right\}
$$

(a) Using max - min composition, find $\mathrm{S}=\mathrm{I}^{\circ} \circ(\mathrm{T} \times \mathrm{O})$.
(b) Using max-product composition, find $\mathrm{S}=1 \mathrm{r} \circ \mathrm{R}$.

Check back soon!
02:34

Problem 8

Relating earthquake intensity to ground acceleration is an imprecise science. Suppose we have a universe of earthquake intensities (on the Mercalli scale), $\mathrm{I}=\{5,6,7,8,9\}$, and a universe of accelerations, $A=\{0.2,0.4,0.6,0.8,1.0,1.2\}$, in $g$. The following fuzzy relation, $R$, exists on the Cartesian space $1 \times \mathrm{A}$ :

$$
\mathrm{R}=\begin{array}{r}
5 \\
6 \\
7 \\
8 \\
9
\end{array}\left[\begin{array}{cccccc}
0.2 & 0.4 & 0.6 & 0.8 & 1.0 & 1.2 \\
0.75 & 1 & 0.85 & 0.5 & 0.2 & 0 \\
0.5 & 0.8 & 1 & 0.7 & 0.3 & 0 \\
0.1 & 0.5 & 0.8 & 1 & 0.7 & 0.1 \\
0 & 0.2 & 0.5 & 0.85 & 1 & 0.6 \\
0 & 0 & 0.2 & 0.5 & 0.9 & 1
\end{array}\right]
$$
If the fuzzy set "intensity about 7 " is defined as

$$
I_7=\left\{\frac{0.1}{5}+\frac{0.6}{6}+\frac{1}{7}+\frac{0.8}{8}+\frac{0.2}{9}\right\}
$$

determine the fuzzy membership of $\frac{1}{7}$ on the universe of accelerations, A.

AG
Ankit Gupta
Numerade Educator

Problem 9

Given the continuous, noninteractive fuzzy sets $\underset{\sim}{A}$ and $\underset{\sim}{B}$ on universes $X$ and $Y$, using Zadeh's notation for continuous fuzzy variables,

$$
\begin{aligned}
& \mathrm{A}=\left\{\int \frac{1-0.1|x|}{x}\right\} \quad \text { for } x \in[0,+10] \\
& \underset{\sim}{B}=\left\{\int \frac{0.2|y|}{y}\right\} \quad \text { for } y \in[0,+5]
\end{aligned}
$$
as seen in Fig. P3.9a:
(a) Construct a fuzzy relation $R$ for the Cartesian product of $\underset{A}{A}$ and B .
(b) Use max-min composition to find ${\underset{B}{B}}^{\prime}$, given the fuzzy singleton $\boldsymbol{A}^{\prime}=\frac{1}{3}$ (see Fig. P3.9b). Hint: You can solve this problem graphically by segregating the Cartesian space into various regions according to the min and max operations, or you can approximate the continuous fuzzy variables as discrete variables and use matrix operations. In any case, sketch the solution.

Check back soon!

Problem 10

Risk assessment of hazardous waste situations requires the assimilation of a great deal of linguistic information. In this context, consider risk as being defined as "consequence of a hazard" multiplied by "possibility of the hazard" (instead of the conventional "probability of hazard"). Consequence is the result of an unintended action on humans, equipment, or facilities, or the environment. Possibility is the estimate of the likelihood that the unintended action will occur. Consequence and possibility are dependent on several factors and therefore cannot be determined with precision. We will use composition, then, to define risk; hence risk $=$ consequence $\circ$ possibility, or

$$
\underline{R}=\mathrm{C} \circ \mathrm{P}
$$
We will consider that the consequence is the logical intersection of the hazard mitigation and the hazard source term (the source term defines the type of initiation, such as a smokestack emitting a toxic gas, or a truck spill of a toxic organic); hence we define consequence $=$ mitigation $\cap$ source term, or

$$
\underset{\sim}{\mathrm{C}}=\underset{\sim}{\mathrm{M}} \cap \underset{\sim}{\mathrm{ST}}
$$

Since humans and their systems are ultimately responsible for preventing or causing non-natural hazards, we define the possibility of a hazard as the logical intersection of human errors and system vulnerabilities; hence possibility $=$ human factors $\cap$ system reliabilities,

$$
\underset{\sim}{\mathrm{P}}=\underset{\sim}{\mathrm{H}} \cap \underset{\sim}{\mathrm{S}}
$$

From these postulates, show that the membership form of the risk, R , is given by the expression

$$
\mu_{\mathbb{R}}(x, y)=\max \left[\min \left(\min \left[\mu_{\mathrm{M}}(x, y), \mu_{\mathrm{ST}}(x, y)\right], \min \left[\mu_{\mathrm{H}}(x, y), \mu_{\mathrm{S}}(x, y)\right]\right)\right]
$$

Check back soon!

Problem 11

A new optical microscope camera uses a lookup table to relate voltage readings (which are related to illuminance) to exposure time. To aid in the creation of this lookup table, we need to determine how much time the camera should expose the pictures at a certain light level. Define a fuzzy set "around 3 volts" on a universe of voltage readings in volts

$$
\underline{V}_{1 \times 5}=\left\{\frac{0.1}{2.98}+\frac{0.3}{2.99}+\frac{0.7}{3}+\frac{0.4}{3.01}+\frac{0.2}{3.02}\right\} \quad \text { (volts) }
$$

and a fuzzy set "around 1/10 second" on a universe of exposure time in seconds

$$
T_{1 \times 6}=\left\{\frac{0.1}{0.05}+\frac{0.3}{0.06}+\frac{0.3}{0.07}+\frac{0.4}{0.08}+\frac{0.5}{0.09}+\frac{0.2}{0.1}\right\} \quad \text { (seconds) }
$$

(a) Find $\underset{\sim}{\mathrm{R}}=\underset{\sim}{\mathrm{V}} \times \mathrm{T}$.

Now define a third universe of "stops." In photography, stops are related to making the picture some degree lighter or darker than the "average" exposed picture. Therefore, let Universe of Stops $=\{-2,-1.5,-1,0, .5,1,1.5,2\}$ (stops). We will define a fuzzy set on this universe as

$$
\underset{Z}{Z}=\text { a little bit lighter }=\left\{\frac{0.1}{0}+\frac{0.7}{0.5}+\frac{0.3}{1}\right\}
$$

(b) Find $\underset{\sim}{\mathrm{S}}=\mathrm{T} \times \underset{\sim}{\mathrm{Z}}$.
(c) Find $\mathrm{M}=\mathrm{R} \circ \underset{\sim}{\mathrm{S}}$ by max-min composition.
(d) Find $\mathrm{M}=\mathrm{R} \circ \mathrm{S}$ by max-product composition.

Check back soon!
01:49

Problem 12

Music is not a precise science. Tactile movements by musicians on various instruments come from years of practice, and such movements are very subjective and imprecise. When a guitar player changes from an A chord to a C chord (major), his or her fingers have to move some distance, which can be measured in terms of frets (e.g., 1 fret $=0.1$ ). This change in finger movement is described by the relation given in the following table. The table is for a six-string guitar: $x_i$ is the string number, for $i=1,2, \ldots, 6$. For example, -0.2 is two frets down and 0.3 is three frets up, where 0 is located at the top of the guitar fingerboard.
The finger positions on the guitar strings for the two chords can be given in terms of the following membership functions:

$$
\begin{aligned}
& \text { C chord }=\left\{\frac{0}{x_6}+\frac{0.3}{x_5}+\frac{0.2}{x_4}+\frac{0}{x_3}+\frac{0.1}{x_2}+\frac{0}{x_1}\right\} \\
& \text { A chord }=\left\{\frac{0}{x_6}+\frac{0}{x_5}+\frac{0.2}{x_4}+\frac{0.2}{x_3}+\frac{0.2}{x_2}+\frac{0}{x_1}\right\}
\end{aligned}
$$

Suppose the placement of fingers on the six strings for a G chord is given as

$$
\mathrm{G} \text { chord }=\left\{\frac{0.3}{x_6}+\frac{0.2}{x_5}+\frac{0}{x_4}+\frac{0}{x_3}+\frac{0}{x x_2}+\frac{0.3}{x_1}\right\}
$$

(a) Find the relation that expresses moving from an A chord to a G chord; call this R .
(b) Use max-product composition to determine CoR.

Adriano Chikande
Adriano Chikande
Numerade Educator
06:38

Problem 13

In neuroscience research, it is often necessary to relate functional information to anatomical information. One source of anatomical information for a subject is a series of magnetic resonance imaging (MRI) pictures, composed of gray-level pixels, of the subject's head. For some applications, it is useful to segment (label) the brain into MRI slices (images along different virtual planes through the brain). This procedure can be difficult to do using gray-level values alone. A standard - or model - brain, combined with a distance criterion, can be used in conjunction with the gray-level information to improve the segmentation process. Define the following elements for the problem:
1. Normalized distance from the model (D)

$$
\underset{\sim}{D}=\left\{\frac{1}{0}+\frac{0.7}{1}+\frac{0.3}{2}\right\}
$$

2. Intensity range for the cerebral cortex ( $\mathrm{L}_{\mathrm{C}}$ )

$$
\mathrm{I}_{\mathrm{C}}=\left\{\frac{0.5}{20}+\frac{1}{30}+\frac{0.6}{40}\right\}
$$

3. Intensity range for the medulla (IM)

$$
\mathrm{I}_{\mathrm{M}}=\left\{\frac{0.7}{20}+\frac{0.9}{30}+\frac{0.4}{40}\right\}
$$

Based on these membership functions, find the following:
(a) $\underset{\sim}{\mathrm{R}}=\mathrm{I}_{\mathrm{C}} \times \underset{\sim}{\mathrm{D}}$
(b) Max-min composition of $I_M \circ \mathrm{R}$
(c) Max-product composition of $\mathrm{I}_{\mathrm{M}} \circ \mathrm{R}$

AG
Ankit Gupta
Numerade Educator

Problem 14

In the field of computer networking there is an imprecise relationship between the level of use of a network communication bandwidth and the latency experienced in peer-to-peer communications. Let $X$ be a fuzzy set of use levels (in terms of the percentage of full bandwidth used) and $\underset{\sim}{Y}$ be a fuzzy set of latencies (in milliseconds) with the following membership functions:

$$
\begin{aligned}
& X=\left\{\frac{0.2}{10}+\frac{0.5}{20}+\frac{0.8}{40}+\frac{1.0}{60}+\frac{0.6}{80}+\frac{0.1}{100}\right\} \\
& \underline{Y}=\left\{\frac{0.3}{0.5}+\frac{0.6}{1}+\frac{0.9}{1.5}+\frac{1.0}{4}+\frac{0.6}{8}+\frac{0.3}{20}\right\}
\end{aligned}
$$
(a) Find the Cartesian product represented by the relation $\mathrm{R}=\underset{\sim}{\mathrm{X}} \times \underset{\sim}{\mathrm{Y}}$.

Now, suppose we have a second fuzzy set of bandwidth usage given by

$$
Z=\left\{\frac{0.3}{10}+\frac{0.6}{20}+\frac{0.7}{40}+\frac{0.9}{60}+\frac{1}{80}+\frac{0.5}{100}\right\}
$$

(b) using max - min composition;
(c) using max-product composition.

Check back soon!

Problem 15

High-speed rail monitoring devices sometimes make use of sensitive sensors to measure the deflection of the earth when a rail car passes. These deflections are measured with respect to some distance from the rail car and, hence, are actually very small angles measured in microradians. Let a universe of deflections be $\mathrm{A}=\{1,2,3,4\}$ where A is the angle in microradians, and let a universe of distances be $\mathrm{D}=\{1,2,5,7\}$ where D is distance in feet. Suppose a relation between these two parameters has been determined as follows:

$$
\left.\mathrm{R}=\begin{array}{c}
\\
\mathrm{A}_1 \\
\mathrm{~A}_2 \\
\mathrm{~A}_3 \\
\mathrm{~A}_4
\end{array} \begin{array}{cccc}
\mathrm{D}_1 & \mathrm{D}_2 & \mathrm{D}_3 & \mathrm{D}_4 \\
1 & 0.3 & 0.1 & 0 \\
0.2 & 1 & 0.3 & 0.1 \\
0 & 0.7 & 1 & 0.2 \\
0 & 0.1 & 0.4 & 1
\end{array}\right]
$$

Now let a universe of rail car weights be $W=\{1,2\}$, where $W$ is the weight in units of 100,000 pounds. Suppose the fuzzy relation of W to A is given by

Using these two relations, find the relation, $\mathrm{R}^{\mathrm{T}} \mathrm{O}=\mathrm{T}$ (note the matrix transposition here)
(a) using max - min composition;
(b) using max-product composition.

Check back soon!

Problem 16

In the field of soil mechanics new research methods involving vision recognition systems are being applied to soil masses to watch individual grains of soil as they translate and rotate under confining pressures. In tracking the motion of the soil particles some problems arise with the vision recognition software. One problem is called "occlusion," whereby a soil particle that is being tracked becomes partially or completely occluded from view by passing behind other soil particles. Occlusion can also occur when a tracked particle is behind a mark on the camera's lens, or the tracked particle is partly out of sight of the camera. In this problem we will consider only occlusions for the first two problems mentioned. Let us define a universe of parameters for particle occlusion, say

$$
\mathrm{X}=\left\{x_1, x_2, x_3\right\}
$$

and a universe of parameters for lens mark occlusion, say

$$
Y=\left\{y_1, y_2, y_3\right\}
$$

Then, define

$$
\mathrm{A}=\left\{\frac{0.1}{x_1}+\frac{0.9}{x_2}+\frac{0.0}{x_3}\right\}
$$
as a specific fuzzy set for a tracked particle behind another particle, and let

$$
\mathrm{B}=\left\{\frac{0}{y_1}+\frac{1}{y_2}+\frac{0}{y_3}\right\}
$$

be a particular fuzzy set for an occlusion behind a lens mark.
Let C be another fuzzy set in which a tracked particle is behind a particle, e.g.,

$$
\mathrm{C}=\left\{\frac{0.3}{x_1}+\frac{1.0}{x_2}+\frac{0.0}{x_3}\right\}
$$

(b) Using max-min composition, find $\underset{\sim}{\mathrm{S}}=\mathrm{C} \circ \mathrm{R}$.

Check back soon!

Problem 17

A common way to control the outlet composition of a distillation column is by controlling the temperature of one of the trays. In order for the system to be controlled properly, the set point of the chosen tray must guarantee that the outlet composition of the distillation column (could be the distillate, the bottoms, or a side stream) is under specifications for the expected range of disturbances. Tray temperature is measured instead of outlet composition because it is hard to measure online the fraction of a compound without dead time (e.g., common gas chromatographs take minutes to make a reading), whereas temperature is easy, fast, and inexpensive to measure. Steady state simulation runs are made for the expected disturbances in order to pick a tray from the distillation column and a temperature value for the control scheme. The problem with this methodology is that there is never a "best" tray temperature for all the disturbances.
For this problem, we define a distillation column (see Fig. P3.17) that separates light hydrocarbons, propane, butane, and iso-butane. We also state that the composition of propane in the bottoms must be $\leq 0.01$ mole fraction, and that the distillation column is made of 20 trays counted from top to bottom.
We define the set $F$ to represent three types of feed flows $=\{$ High flow, Medium flow, Low flow\}. Besides the bulk amount of the feed, another important factor is the comparison in the feed of the key component. We will use propane $\left(\mathrm{C}_3\right)$ as the key component. Define the fuzzy set $\underset{P}{P}=\left\{\right.$ High $\mathrm{C}_3$, Low $\left.\mathrm{C}_3\right\}$. Finally, let us say that from previous raw studies, the prospective trays to be chosen for control are reduced to the set $T=\left\{\right.$ Tray $_{14}$, Tray $_{15}$, Tray $_{16}$, Tray $\left.{ }_{17}\right\}$, where fuzzy set T refers to how well it keeps the column under specifications.

If we have the fuzzy sets
(b) $\underset{P}{ }$ and T are highly related. If we have

$$
\left.\mathrm{T}=\begin{array}{cccc}
\text { Tray }_{14} & \text { Tray }_{15} & \text { Tray }_{16} & \text { Tray }_{17} \\
0.1 & 0.3 & 0.8 & 0.7
\end{array}\right]
$$

then find the Cartesian product $\mathrm{S}=\underline{\mathrm{P}}^{\mathrm{T}} \times \mathrm{T}$.
(c) Since outlet composition is not easily measured, it is still important to know how F and T are related. This knowledge can be acquired with a max-min composition; find $\mathrm{C}=\mathrm{R} \circ \mathrm{S}$.

Check back soon!

Problem 18

Pyrolysis is a widely used high-temperature reaction to produce ethylene and propylene. When this product leaves the furnace reactor it is necessary to reduce their temperature as quickly as possible in order to stop the reaction and avoid producing undesired products. This "quenching" of the products is made in equipment that works and looks just like any heat exchanger, with the difference that a reaction is actually happening in the tubes. Very good heat transfer is obtained when the tubes are clean, but coke deposition on the inner surface of the tubes reduces the heat transfer coefficient by a considerable amount and it also increases the pressure drop. Therefore, coke deposition in the pyrolysis of light hydrocarbons becomes an important factor in the design of the equipment (usually, the equipment needs cleaning every four months). An experiment was set in order to determine the coke deposition in the exchangers built with 10 tubes for different conditions. The different conditions were:

$$
\begin{aligned}
& X_1=\text { low tube diameter } \\
& X_2=\text { big tube diameter } \\
& X_3=\text { tubes made of material } 1 \\
& X_4=\text { tubes made of material } 2 \\
& X_5=\text { high pressure } \\
& X_6=\text { very high pressure }
\end{aligned}
$$

For every run, all of the 10 tubes were examined and distributed in four categories by percentage. Not all the tubes were expected to get exactly the same amount of coke deposition because there is never perfect distribution of the feed into the 10 tubes.
The examination categories are: High deposition, Med-High deposition, Med deposition, and Moderate deposition. The results are as follows:
(a) It is desired to find the similarity among the six pyrolysis conditions; use the max-min method to find the similarity.
(b) Use the cosine amplitude method to find the similarity.

Check back soon!
03:45

Problem 19

A structural designer is considering four different kinds of structural beams $\left(S_1, \ldots, S_4\right)$ for a new building. Laboratory experiments on the deflection resistance for these four different kinds of beams have been performed, and the engineer wants to determine their suitability in the new structure. The following data have been observed, based on the overall deflection capacity of each bean type:
Using the cosine amplitude method determine the similarity of the four beam types.

Nick Johnson
Nick Johnson
Numerade Educator

Problem 20

Given a particular stream gauge record from three gauging stations, $g_1-g_3$, during major rainstorms, we can count the years that have had storms with equal magnitude over different periods to get a quantity used in the design of storm sewers: the number of years with similar storms over a specified time period.
Find the similarity relation, $R$, among the three gauges using the max-min method.

Check back soon!
02:13

Problem 21

A certain type of virus attacks cells of human body, as explained in Example 3.7. The infected cells can be visualized using a special microscope; the virus causes the infected cells to have a black spot in the image as depicted in Fig. 3.6. The microscope generates digital images that medical doctors can analyze to identify the infected cells. A physician analyzes different samples of an infected organ (S) and classifies it as Not infected, Moderately infected, and Seriously infected. Five samples were analyzed as given in the table below:

Mikayla Stephens
Mikayla Stephens
Numerade Educator

Problem 22

In the statistical characterization of fractured reservoirs, the goal is to classify the geology according to different kinds of fractures, which are mainly tectonic and regional fractures. The purpose of this classification is to do critical simulation based on well data, seismic data, and fracture pattern. After pattern recognition (using Cauchy-Euler detection algorithms or other methods) and classification of the fracture images derived from the outcrops of fractured reservoirs, a geological engineer can get different patterns corresponding to different fracture morphologies. Suppose the engineer obtains five images ( $I_1 \ldots I_5$ ) from five different outcrops of fractured reservoirs, and their percentage values corresponding to three kinds of fractures (Tectonic fracture, Regional fracture, and Other fracture), as given below:
Develop a similarity relation
(a) using the cosine amplitude method and
(b) using the max-min method.
(c) Since the similarity relation will be a tolerance relation, fund the associated equivalence relation.

Check back soon!
13:01

Problem 23

The latitude of the receiver can affect the positional accuracies of geo-positioning-system (GPS) results because the GPS was designed primarily for operation close to the equator. The accuracy of a GPS degrades as the distance from the equator increases. We define the following fuzzy sets:
discrete set for position accuracies (P)
discrete set for latitude ( L )
(a) Discuss the physical meaning of the relation $\underline{R}_1=\underline{\mathrm{P}} \times \underline{\mathrm{L}}$.

Position accuracies of GPS are also affected by ionospheric effects. Ionospheric effects are strongest at the magnetic equator and they weaken as the GPS gets closer to the poles. We define a fuzzy discrete set for ionospheric effect (I).
(b) Discuss the physical meaning of the relation $\mathrm{R}_2=\mathrm{L} \times \underset{\sim}{\mathrm{I}}$.
(c) Through composition we can relate the ionospheric effect with the position accuracies, i.e., by calculating $\underset{\sim}{\mathrm{IP}}=\mathrm{R}_1 \circ \mathrm{R}_2$. Discuss the physical meaning of this composition.
(d) Five sets of GPS measurements were made and the quality of each range to a single satellite was graded based on its elevation angle, in degrees. The quality ranged from poor to acceptable and to high quality. The following table shows the distribution of GPS ranges for each of the five sets of measurements. Using the max-min similarity method find the fuzzy similarity relation for the five sets.
(e) Find the equivalence relation of the matrix developed in part (d).

Sarah Lewites
Sarah Lewites
Numerade Educator

Problem 24

Over the last decade Calgary, Alberta, has made use of a large number of storm-water ponds to mitigate flooding during extreme rainfall events by reducing the peak flows in trunk sewers or receiving waterways. The ponds have control structures that have data recorders monitoring water levels during the storm events. More ponds are built as urban development occurs. To determine the similarity of the pond performances within the city quadrants would be useful in determining if citywide drainage policies work effectively. Let the regions of the city be represented by the four quadrants (NW, NE, SW, SE). The ponds in the area can have performance based on design capacities broken down into three categories of stormwater filling levels: low, moderate, and high values based on expectations and percentage of ponds in the category. The following table represents a recent storm event summary:
(a) Use the cosine amplitude method to express these data as a fuzzy relation.
(b) Comment on the similarity of the ponds in the four quadrants of the city.
(c) Find the equivalence relation associated with the matrix calculated in part (b). How does the equivalence relation differ, in terms of the physical significance, from the original tolerance relation?

Check back soon!
01:51

Problem 25

The accompanying Sagittal diagrams (Fig. P3.25) show two relations on the universe, $X=$ $(1,2,3\}$. Are these relations equivalence relations [Gill, 1976]?

Chris Trentman
Chris Trentman
Numerade Educator
04:04

Problem 26

For Example 3.6 in this chapter, recalculate the fuzzy relation $T$ using
(i) Equation (3.25)
(ii) Equation (3.26)
(iii) Equation (3.27)
(iv) Equation (3.28)
(v) Equation (3.29), where

$$
f(\bullet)- \begin{cases}1-e^{-x}, & \text { for } x \geq 0 \\ e^x-1, & \text { for } x \leq 0\end{cases}
$$

Suman Saurav Thakur
Suman Saurav Thakur
Numerade Educator

Problem 27

Fill in the following table using Eqs. (3.25)-(3.29) to determine values of the composition $\underset{\sim}{\mathrm{B}}=\underset{\sim}{\mathrm{A}} \circ \mathrm{R}$ for the fuzzy relation

$$
\underline{\mathrm{R}}=\begin{gathered}
x_1 \\
x_2 \\
x_3
\end{gathered}\left[\begin{array}{ccc}
y_1 & y_2 & y_3 \\
0.1 & 0.2 & 0.3 \\
0.4 & 0.5 & 0.6 \\
0.7 & 0.8 & 0.9
\end{array}\right]
$$

Comment on the similarities and dissimilarities of the various composition methods with respect to the various antecedents, A .

Check back soon!