Professional Documents
Culture Documents
Production Examples
Production Examples
Production Examples
On each example worksheet, read the comments at the bottom of the sheet, then
click Tools Solver... to examine the decision variables, constraints, and objective.
To find the optimal solution, click the Solve button.
One of the areas where linear and non-linear programming is applied most frequently is
in manufacturing and production. There are many different ways in which the Solver can
be used to increase productivity, lower cost, reduce waste, etc. We will limit ourselves to
5 different types of models.
First, in the ProductMix worksheet we examine a production mix model. Here we see how
to use parts to build different products and maximize profit.
Second, we look at a machine allocation problem, in two different versions (Alloc1 and
Alloc2). This model determines which machines to use to produce products, in order to
meet a certain demand and minimize cost.
Third, in worksheets Blend1 and Blend2, we look at a 'continuous' rather than 'discrete'
production problem, where the end product requires certain qualities and is a mixture
of previously produced products. This kind of blending problem is very common in
the oil industry and in agriculture, for example.
Fourth, worksheet Process is a process selection problem, where we have several
different ways of producing something (planed wood in this example) and we want to
pick the process that minimizes cost (or maximizes profit).
Finally, in worksheet Cutstock we look at a classic example of a cutting stock problem.
In this model we determine how to cut steel sheets so as to minimize the waste of steel.
254879305.xls.ms_office
Part Name
TV set
Stereo
Speaker
1
1
2
1
2
1
0
2
1
1
0
0
1
0
1
Number to Build->
Inventory
No. Used
Chassis
Picture Tube
Speaker Cone
Power Supply
Electronics
450
250
800
450
600
0
0
0
0
0
Profits:
By Product
Total
$0
$0
$0
$0
Problem
Your company builds TVs, stereos and speakers, using a common parts inventory of power supplies,
speaker cones, etc. Parts are in limited supply. What is the best combination of products to build
that maximizes profit?
Solution
1) The variables are clearly the number of TVs, stereos and speakers to build. In this worksheet, they
are given the name Number_to_build.
2) The constraints specify that the number of parts used cannot exceed the supply. This leads to
Number_used <= Number_available
There is also the logical constraint
Number_to_build >= 0 via the Assume Non-Negative option
3) The objective is to maximize profit. In the ProductMix worksheet this is defined as Total_profit.
Remarks
Although this is a good example of a product mix problem, bear in mind the limitations of the
model. For example, market demand and price elasticity is not included in the model -- we assume
that it doesn't matter how many TVs we build, we will always be able to sell them. Nor are there any
pre-specified minimum or maximum of products that need to be made. The effect of introducing
these restrictions can be studied by examining a Sensitivity Report, which you can select from the
dialog appears with the message 'Solver found a solution.'
Page 2
254879305.xls.ms_office
$200
$275
$325
Alpha-1000
Alpha-2000
Alpha-3000
$1.50
$1.80
$1.90
Products
per day
(Max)
Number of
machines
40
60
85
8
5
3
1
1
1
Alpha-1000
Alpha-2000
Alpha-3000
300
300
300
900
750
Alpha-1000
Alpha-2000
Alpha-3000
Total
Demand
40
60
85
Alpha-1000
Alpha-2000
Alpha-3000
Cost
$2,360.00
Problem
A company has three different types of machines that all make the same product. Each
machine has a different capacity, start-up cost and cost per product. How should the company
produce its product with the available machines to meet the daily demand?
Solution
1) The variables are the number of machines to use and the number of products to make on
each machine. In worksheet Alloc1, these are given the names Products_made and
Machines_used.
2) First, there are the logical constraints. These are
Products_made >= 0 via the Assume Non-Negative option
Machines_used >= 0 via the Assume Non-Negative option
Machines_used = integer.
Second, there are the demand and capacity constraints. These are:
Machines_used <= Machines_available
Products_made <= Maximum_products
Total_made >= Demand
3) The objective is to minimize cost. This is defined on the worksheet as Total_cost.
Remarks
Notice that we used an integer constraint to make sure no fractions of machines were used.
This has the usual drawback; the problem is much more difficult to solve than the 'relaxed'
version without the integer constraint. When large numbers of machines
are involved, the
Page 3
integer constraint can often be dropped. It is most often not critical whether 1586 or 1587
machines are used, for example. This means that a number of 1586.4 would be acceptable.
254879305.xls.ms_office
Alpha-1000
Alpha-2000
Alpha-3000
Additional
cost per
product
Products per
day (Max)
Number of
machines
$200
$275
$325
$1.00
$1.80
$1.90
40
60
85
8
5
3
Monday
Tuesday
Wednesday
Thursday
Friday
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Monday
Tuesday
Wednesday
Thursday
Friday
0
0
0
0
0
0
600
0
0
0
0
-600
-600
800
0
0
0
0
-1400
-1400
1000
0
0
0
0
-2400
-2400
725
0
0
0
0
-3125
-3125
750
Monday
Tuesday
Wednesday
Thursday
Friday
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Total
Cost
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
Problem
A company has three different types of machines that all make the same product. Each machine has
a different capacity, start-up cost and cost per product. How should the company produce its
product with the available machines to meet the demand over a week's time?
Solution
The solution is very similar in structure to the one found on worksheet Alloc1.
1) The variables are the number of machines to use and the number of products to make on each
machine. In worksheet Alloc2, these given the names Products_made and Machines_used.
2) First, there are the logical constraints. These are
Products_made >= 0 via the Assume Non-Negative option
Machines_used >= 0 via the Assume Non-Negative option
Machines_used = integer.
Second, there are the demand and capacity constraints. These are:
Alpha1000s_used <= Alpha1000s_available
Alpha2000s_used <= Alpha2000s_available
Alpha3000s_used <= Alpha3000s_available
Products_made <= Maximum_products
Page 4
254879305.xls.ms_office
Magnesium
contents
(relative to
required
quality)
2.3
1.6
1.2
4.1
2000
2500
1300
3000
3.5
4
4
2
Year 2
0
0
0
0
0
Year 3
0
0
0
0
0
Year 4
0
0
0
0
0
Year 5
0
0
0
0
0
Year 2
0
0
0
0
0
3100
Year 3
0
0
0
0
0
3500
Year 4
0
0
0
0
0
3700
Year 5
0
0
0
0
0
4000
0.7
1.5
0.7
Quarry 2
Quarry 3
Quarry 4
Cost to
Maximum keep quarry
production
open per
per year
year
(tons)
($Million)
Year 1
0
0
0
0
0
Year 1
0
0
0
0
0
4500
Year 1
0
0
0
0
Year 2
0
0
0
0
Year 3
0
0
0
0
Year 4
0
0
0
0
Year 5
0
0
0
0
Year 1
Year 2
Year 3
Year 4
Year 5
Total Amount of
Calcium
Total Amount
Required
0.9
1.2
1.1
0.8
Year 1
Year 2
Year 3
Year 4
Year 5
Quarry 1
Quarry 2
Quarry 3
Quarry 4
Calcium restrictions
Calcium Required
per Ton (Minimum)
Magnesium restrictions
Total Amount of
Magnesium
Page 5
254879305.xls.ms_office
Total Amount
Required
Magnesium Required
per Ton (Minimum)
Cost
1.9
1.7
2.8
1.9
2.1
Year 1
$0.00
Year 2
$0.00
Year 3
$0.00
Year 4
$0.00
Year 5
$0.00
Total
$0.00
Problem
A company owns four rock quarries from which it can extract limestone with different qualities. Two
qualities are important, the relative amount of calcium and magnesium in the stone. The company
must produce a certain total amount of limestone, with certain qualities, each year. There is a large
fixed cost to keep a quarry operating for extraction purposes each year. Which quarries should be
used each year, and how much limestone should each one produce each year?
Solution
The solution is very similar in structure to the one found in worksheet Blend1.
1) The variables are 0-1 or binary integer variables which determine whether each quarry is open,
and amounts of limestone to be extracted from each quarry. These variables occur in each year.
In worksheet Blend2, these variables the names Quarry_decisions and Amounts_produced.
2) First, there are the logical constraints. These are
Amounts_produced >= 0 via the Assume Non-Negative option
Quarry_decisions = binary
Second, there are contraints on the total production and the amount that can be produced at each
quarry. These constraints are:
Total_produced >= Total_required
Amounts_produced <= Maximum_Production
The right hand side of the second constraint depends on the binary integer variables.
Third, there are constraints on the quality (calcium and magnesium content) of the limestone:
Calcium_production >= Calcium_requirement
Magnesium_production >= Magnesium_requirement
Both the left-hand and right-hand sides of these constraints depend on the Amounts_produced
decision variables.
Fourth, there is a constraint that limits the number of quarries that can be open each year:
Number_of_open_quarries <= 3
3) The objective is to minimize the cost of operating the quarries. This is defined on the worksheet as
Total_cost.
Remarks
See the comments on worksheet Blend1 about the characteristics of blending problems, which also
apply here.
Page 6
254879305.xls.ms_office
Process Selection
A planing mill uses 3 different types of planers. What planers should the company use
to minimize cost? The total job has to be finished in
hours.
Characteristics of planers
Planer 1
Planer 2
Planer 3
Speed
(ft/min)
Cost
($/hour)
Maximum
wood
thickness
(inches)
5
7
8
$150
$190
$225
6
4
2
3"
0
0
5"
0
0
600
0
300
Hours
Cost
0.00
0.00
0.00
$0.00
$0.00
$0.00
$0.00
Problem
A planing mill has three different planers. Each planer has a different speed, cost to operate and
maximum thickness of wood it can handle. What planers should the mill use to minimize cost, given
an amount of wood and no more than 3 hours to do the job?
Solution
The solution is structurally very similar to the one found on worksheet Alloc1.
1) The variables are the amounts of wood that go through the different planers. In worksheet
Process, these are given the names Wood_through_planer1, Wood_through_planer2 and
Wood_through_planer3.
2) The logical constraints are all defined via the Assume Non-Negative option:
Wood_through_planer1 >= 0
Wood_through_planer2 >= 0
Wood_through_planer3 >= 0
The time and demand constraints give
Total_hours <= Hours_available
Total_planed >= Demand
3) The objective is to minimize cost and this is defined on the worksheet as Total_cost.
Remarks
This is only a small example of a process selection. An example where process selection is very
important is the oil industry. A process selection model is often used to decide what method to use
to create a product.
Page 7
254879305.xls.ms_office
Cutting Stock
A steel mill produces sheets of steel in 3 sizes. These sizes are 100, 80 and 55 inches. Unfortunately,
demand is in 3 other sizes; 45,30 and 18 inches.How should the mill cut the sheets to minimize waste?
Possible combinations
45" sheet
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
100" sheet
80" sheet
55" sheet
Totals
Demand
30" sheet
18" sheet
2
1
1
0
0
0
0
1
1
0
0
0
1
0
0
0
1
0
3
2
1
0
1
0
2
1
0
0
1
0
0
1
3
0
2
3
5
0
1
1
2
4
0
1
3
7
150
12
200
26
175
Waste
(inches)
Number of
sheets
Total
Waste
10
7
1
10
4
16
10
5
17
2
14
8
10
7
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
10
7
1
10
4
16
10
5
17
2
14
8
10
7
1
Total
122
Problem
A steel mill produces sheets of steel in three different sizes. Demand, however, is in 3 other, smaller, sizes.
How should the company cut the sheets of steel in order to minimize waste?
Solution
1) There are only a limited number of ways to cut the sheets. The variables are the number of times we have
to cut a sheet in a certain way. In worksheet Cutstock these are defined as Sheets_used.
2) The constraints are simple and straightforward.
Sheets_made = Demand
Sheets_used >= 0 via the Assume Non-Negative option
Sheets_used = integer
3) The objective is to minimize waste. This is defined on the worksheet as Total_waste.
Remarks
In some situations it may seem rather difficult to write out all the possibilities for cutting stock as is done in
this model. There is a technique that lets the computer do this, called column generation. It is beyond the
scope of this example to fully discuss this technique.
Page 8