Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

ISyE 3133, Tutorial 6

with Partial/Selected Answers


Learning Goals
1. Further IP modeling experience
2. Fixed charge, semi-continuous variables

Exercise 1. A power plant has four boilers. If a given boiler is operated, it can be used to
produce a quantity of steam (in tons) between the minimum and maximum given in Table 1.
The cost of producing a ton of steam on each boiler is also given, as well as the fixed cost of
operating each boiler. Steam from the boilers is used to produce power on three turbines. If
operated, each turbine can process an amount of steam (in tons) between the minimum and
maximum given in Table 2. The cost of processing a ton of steam and the power produced by
each turbine is also given. The power plant must produce at least 9,000 Kwh of power.

Boiler # Min. Max. Cost per ton ($) Fixed cost ($)
1 400 900 9 160
2 500 700 7 200
3 300 600 8 190
4 240 800 6 250

Table 1: Data for each of the boilers

Turbine # Min. Max. Kwh per ton of steam Processing Cost per Ton ($)
1 100 700 7 9
2 400 800 3 4
3 300 600 6 6

Table 2: Data for each of the turbines

(a) Formulate a linear MIP that can be used to minimize the cost while satisfying all con-
straints. Clearly specify the decision variables (and what they mean), as well as the
objective function and constraints.
Parameters:
fi = fixed cost of boiler i
cbi = cost per ton of steam produced by boiler i
lib = minimum steam production for boiler i if operated
ubi = maximum steam production for boiler i if operated
pj = Kwh output per ton of steam processed by turbine j
ctj = cost per ton of steam processed by turbine j
ljt = minimum steam processed by turbine j if operated
utj = maximum steam processed by turbine j if operated
Variables:
yi = 1 if boiler i operated

1
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
xi = tons of steam produced by boiler i
wj = 1 if turbine j used
vj = tons of steam used by turbine j,
i = 1, . . . , m. j = 1, . . . , n, where there are m boilers and n turbines. An alternative to
using both x and v variables would be to use xij = tons of steam produced by boiler i to
be used in turbine j. Objective is
m
X n
X
min (fi yi + cbi xi ) + ctj vj .
i=1 j=1

Total steam produced by boilers must match total steam processed at turbines (assuming
all boilers can supply all turbines, and no steam “dumping” is permitted):
m
X n
X
xi = vj .
i=1 j=1

Adequate power must be generated:


n
X
pj vj ≥ 9000.
j=1

Variable bounds on steam generated by boilers:

lib yi ≤ xi ≤ ubi yi , ∀i = 1, . . . , m.

Variable bounds on steam processed by turbines:

ljt wj ≤ vj ≤ utj wj , ∀j = 1, . . . , n.

(b) Write linear constraints that correctly model the restriction that if both boilers 1 and 2
operate, then boilers 3 and 4 must not operate. (You may introduce new variables if you
need to.) x3 ≤ 2 − (x1 + x2 ) and x4 ≤ 2 − (x1 + x2 )
(c) Suppose that if we operate both boilers 3 and 4, then instead of the fixed cost being 440,
it drops to 340 since they are able to share some resources. How would you modify your
linear IP formulation to reflect this cost change? (You may introduce new variables if you
need to.) Let s be a new binary variable that must be zero if either (or both) boilers 3 or
4 do not operate. Then in the objective we subtract 100s, and in the constraints s ≤ x3
and s ≤ x4 .

Exercise 2. Jack and Jill love to watch DVDs and, as a couple, have vowed to always watch
their DVDs together as a form of spending time with each other. They have subscribed for a 1
month free trial and estimated that they will watch 20 DVDs by the end of the free trial. They
have pre-selected 100 DVDs that they would like to watch and now want to decide which 20
out of these 100 to watch during the free trial month. These 100 DVDs are divided as follows:
DVDs 1-20 are action movies, DVDs 21-40 are romantic comedies, DVDs 41-60 are TV series,
DVDs 61-80 are documentaries and DVDs 81-100 are science fiction.

2
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
They associated an estimated common satisfaction index si to DVD i, for each DVD i =
1, . . . , 100, and wish to maximize their total estimated common satisfaction during the free trial
month. They also established the following constraints so that each of them will not be too
unhappy: they must choose at least 3 action movies and at least 3 romantic comedies, but not
more than 5 science fiction and no more than 2 documentaries.
(Assume that they will be able to obtain all movies they request.)
a) Write an integer linear programming formulation that Jack and Jill can use to solve
this problem of maximizing their estimated common satisfaction. Clearly indicate your
decision variables and what they mean.
Let binary decision variable xi = 1 indicate that Jack and Jill chose DVD i to watch
during the free trial month; xi = 0 otherwise. The model is as follows.
100
X
max si xi
i=1
100
X
s.t. xi = 20, (will watch 20 DVDs)
i=1
20
X
xi ≥ 3, (watch at least 3 action movies)
i=1
X40
xi ≥ 3, (watch at least 3 rom-coms)
i=21
X80
xi ≤ 2, (watch at most 2 documentaries)
i=61
X100
xi ≤ 5, (watch at most 5 sci-fi movies)
i=81
xi ∈ {0, 1}, ∀i = 1, . . . , 100.

b) Write one or more linear constraints to express the following additional conditions: (you
may add decision variables if necessary, as long as you make it clear what they mean.
Also, if you use a Big-M constant, explicitly state its value and how you got to it)
i. Since DVDs 11, 12 and 13 are part of a series, either they get all of them, or they
get none of them.
The easiest approach is just to notice that the three corresponding variables must
have the same values: x11 = x12 and x12 = x13 suffice.
ii. DVDs 41–50 are respectively seasons 1–10 of Friends, so they have decided that if
they watch one season of Friends, they must watch all seasons before it. For example,
if they watch season 6, they must also watch seasons 1–5.
The easiest approach is just to notice that these variables must be non-increasing,
so if xi−1 = 0 then xi = 0 for each i = 42, . . . , 50. This can be modeled linearly as

xi ≤ xi−1 , ∀i = 42, . . . , 50.

iii. If they watch at least 2 DVDs out of 24, 34 and 77, then they must not watch more
than one DVD out of 27, 35 and 79.

3
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
Since the variables are binary, this can be written as “if x24 + x34 + x77 − 1 > 0 then
x27 + x35 + x79 − 1 ≤ 0” and modeled using the formula:

x24 + x34 + x77 − 1 ≤ 2y and x27 + x35 + x79 − 1 ≤ 2(1 − y),

where y is an additional binary variable.


c) Jack and Jill realized that their satisfaction model for DVDs 84 and 85 is slightly different.
If they watch exactly one DVD among 84 and 85, then the satisfaction index is t. Else it
is zero. How would you model this fact using only linear constraints and linear objective
function? (You may add auxiliary variables.)
This question is saying: x84 + x85 = 0 means satisfaction from these two DVDs is 0,
x84 + x85 = 1 means satisfaction from these two DVDs is t, while x84 + x85 = 2 means
satisfaction from these two DVDs is s84 + s85 . We can see that this can be modeled as

tx84 + tx85 + (s84 + s85 − 2t)z,

where z = 1 if x84 + x85 = 2 and z = 0 otherwise. This is the same as saying that z = 1
if x84 = 1 and x85 = 1, and z = 0 otherwise, or, equivalently, that z = x84 x85 . This can
be modeled linearly as

z ≤ x84 , z ≤ x85 , and z ≥ x84 + x85 − 1.

Extra Exercises to be done Outside Class Time for Practice


Extra Exercises 1. Model the following logical statements with linear constraints, where x,
y and z are binary variables.
(a) If x + y ≥ 1 then z = 1.
(b) If x = 0 and y = 0 then z = 0.
(c) z = x or y. In other words, x, y and z satisfy the following “truth table”.
x y z
0 0 0
1 0 1
0 1 1
1 1 1
What relationship do you notice between the resulting linear constraints and those you
found for the first two questions?
(d) If x = 1 and y = 1 then z = 1.
(e) If x = 0 or y = 0 then z = 0.
(f) z = x and y. In other words, x, y and z satisfy the following “truth table”.
x y z
0 0 0
1 0 0
0 1 0
1 1 1

4
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
What relationship do you notice between the resulting linear constraints and those you
found for the previous two questions?
[Note: you should see that (a) and (b) together are equivalent to the statement z = x or y,
where variable values of 0 and 1 are interpreted as “false” and “true” respectively. Also, (d)
and (e) together are equivalent to the statement z = x and y.

Extra Exercises 2. Write linear constraints to model the following logical statements, where
x1 , x2 , x3 , x4 are binary variables. You may introduce additional variables if needed.
(a) x4 = x1 x2 x3 .
(b) x4 = (x1 and x2 ) or x3

Extra Exercises 3. There are n supply stations and m demand centers. Let i = 1, ..., n
represent the supply stations and let j = 1, ..., m be the demand centers. Each supply station
has a capacity of at most si tons of products and each demand center requires at least dj tons
of products. (You can model the amount of product as a continuous variable.) The products
are shipped by using trucks. Each truck can carry at most b tons of products. The cost of
sending a truck from supply station i to demand center j is cij , regardless of the amount of
product carried by the truck. Note that (1) Each truck travels from exactly one supply station
to exactly one demand node. (2) Any number of trucks can be sent from supply station i to
demand center j. Write a mixed integer linear program to minimize the total cost.

Extra Exercises 4. (Based on Winston & Albright, Practical Management Science, Example
3.3) The Pigskin Company produces footballs. Pigskin must decide how many footballs to
produce each month. It has decided to use a 6-month planning horizon. The forecasted demands
for the next 6 months are 10,000, 15,000, 30,000, 35,000, 25,000 and 10,000. Pigskin wants to
meet these demands on time, knowing that it currently has 5,000 footballs in inventory and
that it can use a given months production to help meet the demand for that month. During
each month there is enough production capacity to produce up to 30,000 footballs, and there is
enough storage capacity to store up to 10,000 footballs at the end of the month, after demand
has occurred. The forecasted production costs per football for the next 6 months are $12.50,
$12.55, $12.70, $12.80, $12.85, and $12.95, respectively. The holding cost per football held
in inventory at the end of the month is figured at 5% of the production cost for that month.
The selling price for footballs is not considered relevant to the production decision because
Pigskin will satisfy all customer demand exactly when it occurs at whatever the selling price
is. Therefore, Pigskin wants to determine the production schedule that minimizes the total
production and holding costs.
(a) Formulate a linear programming model to help Pigskin plan their next 6 months produc-
tion.
(b) Assume now that in addition to the given production and holding costs, there is a fixed
cost of $5000 during any month in which there is positive production. Assume now that
there is a storage capacity of 20000 footballs. Modify your linear programming model, by
including integer variables and linear constraints, to develop a mixed integer programming
model to help Pigskin plan their next 6 months production.

Extra Exercises 5. Hunter Auto Parts (HAP) is an Australian manufacturer of specialized


shock absorbers for modified Japanese imports. HAP estimates that the demand for these
shock absorbers will be 5,000 for the first quarter, 6,500 for the second quarter, 10,000 for the
third quarter and 7,000 for the fourth quarter. HAP estimate that due to both the fluctuating

5
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
Australian dollar and seasonal traffic in Japanese imports that each shock absorber will cost
them $50, $80, $40, and $70 to manufacture during each of the four quarters respectively. Shock
absorbers remaining at the end of each quarter incur storage costs of $20 per unit. HAP does
not require any inventory to remaining at the end of the year, but is also not averse to having
some on hand if this occurs.
(a) Formulate a linear programming model to help HAP plan their next year’s production so
as to minimize the total manufacturing and inventory costs.
(b) Now consider a minimum batch size for the manufacturing operation. The costs HAP
estimated for the operation are predicated on an efficient number of units manufactured
in the quarter, estimated by the factory production engineer to be around 15,000 shock
absorbers. Develop a mixed integer linear programming model to determine how HAP
can minimize the cost of meeting (on time) the demands of the next year. [Important
note: your model should permit the possibility that zero shock absorbers are produced in
a quarter. The model should require that if some production occurs, at least 15,000 must
be made, but it should be possible that none are made.]

Extra Exercises 6. A machine tool plant owns four different machines on which it can process
jobs. A job may take different times if processed on different machines. A job cannot be divided
between machines, that is, each job must be processed by exactly one machine. Each machine
can process more than one job. If a machine is used at all, then a setup time is needed. The
relevant times for four jobs are given, in minutes, in the following table.

Job 1 Job 2 Job 3 Job 4 Machine Setup


Machine 1 42 70 930 710 10
Machine 2 340 43 120 7 20
Machine 3 560 32 40 9 60
Machine 4 71 760 5 80 85

6
This study source was downloaded by 100000786859274 from CourseHero.com on 04-14-2024 22:07:19 GMT -05:00

https://www.coursehero.com/file/30771168/Tutorial6withAnswerstoExercises1and2pdf/
Powered by TCPDF (www.tcpdf.org)

You might also like