Linear programming: problem description

A company produces four different (A, B, C, D) products

from three different resources (R₁, R₂, R₃).
The resources are Products
bounded by their Resources Capacities
A B C D
capacities. The
R₁ 1 0 2 1 280
capacities of R₂ and
R₂ 2 1 0 0 140
R₃ should be used
R₃ 0 1 1 1 120
completely. From the
products A and B in Product prices 4 5 6 8
total we need to
From the product B we need to produce
produces as much as
at least five unit more than from D.
from C.
Maximize the profit !
3/10/2010 Tibor Illés – Optimisation – 2009 1
 Linear programming: problem description
In a factory using three machines (M₁, M₂, M₃) five different
products (A₁, A₂, A₃, A₄, A₅) can be produced. All products are
manufactured using all three machines. The processing time of
different products are different on given machines. Specific
processing times for different machines, theircapacities in
working hours are shown on the following tables. The precies of
different products are 2, 3, 2, 4, 2, respectively.
Machines Products Capacities
Maximize the A₁ A₂ A₃ A₄ A₅
profit under the
M₁ 1 2 4 3 2 480
following
M₂ 3 1 1 5 3 460
constraints:
M₃ 2 3 1 1 5 450

3/10/2010 Tibor Illés – Optimisation – 2009 2

 A simplified blending problem

An oil refinery produces four types of raw gasoline:

alkylate, catalytic-cracked, straight-run and isopentane.
Two important characteristics of each gasoline are its
performance number PN (indicating antiknock properties) and its
vapor pressure RVP (indicating volatility). These two
characteristics, together with the production levels in barrels per
day, are as follows:
PN RVP Barrels produced These gasolines can be sold
Alkylate 107 5 3814 either raw at £27.87 per
Catalytic-cracked 93 8 2666 barrel, or blended into
Straight-run 87 4 4016 aviation gasoline (Avgas A
Isopentane 108 21 1300 and/or Avgas B).

3/10/2010 Tibor Illés – Optimisation - 2008 3

 A simplified blending problem: continue
Quality standards impose certain requirements on the aviation
gasolines: these requirements, together with the selling prieces,
are as follows:
PN RVP Price per barrel
Avgas A at least 100 at most 7 £42.13
Avgas B at least 91 at most 7 £38.47

The refinery aims for the production plan that yields the largest
possible profit. Formulate as a linear programming problem (in
standard form).

Remark: The PN and RVP of each mixture are simply weighted

averages of the PNs and RVPs of its constituents.
3/10/2010 Tibor Illés – Optimisation - 2008 4
 Resource allocation problem
Ajax Ltd. manufactures and sells three types of computer: α-PC,
β-NB and γ-WS. For the moment we assume that all production
during the week can and will be sold immediately.
α-PC β-NB γ-WS Net profit equals the sales
Net profit 160 210 310 in £s price of each computer
Labour 10 15 20 in hours minus the direct cost of
purchasing components, producing computer cases, and
assembling and testing computer. This week, 120 hours are
available on the A-line test equipment where assembled α-PCs
and β-NBs are tested, and 48 hours are available on the C-line test
equipment where assembled γ-WSs are tested. The testing of
each computer takes 1 hour. In addition, production is constrained
by the availability of 2000 labor hours for product assembly.
3/10/2010 Tibor Illés – Optimisation - 2008 5
 Resource allocation problem: dual version

The Immense Computer Co. is experiencing a rapid growth in sales.

As a result, Immense has insufficient capacity for testing and assembling their
computers, and the director of purchasing is looking to rent capacity from other
smaller, companies. She is considering approaching Ajax to offer to rent its
capacity on a weekly basis.
In particular, she wishes to determine nonnegative prices per hour
of A-line test capacity, nonnegative prices per hour of C-line test
capacity and nonnegative prices per hour of labour capacity to offer
Ajax that will induce Ajax into agreeing to rent its resources rather
than to use them in the manufacture of its own products. At the
same time she wishes to pay the least amount for these resources. A
number of large manufacturing firms are concered with the economic viability of
their small suppliers.

3/10/2010 Tibor Illés – Optimisation - 2008 6

 Resource allocation problem: dual version

The Immense Computer Co. is experiencing a rapid growth in sales.

As a result, Immense has insufficient capacity for testing and assembling their
computers, and the director of purchasing is looking to rent capacity from other
smaller, companies.α-PC β-NB approaching
She is considering γ-WS Ajax to offer Capacity
to rent its
capacity
Net on a weekly160
profit basis. 210 310 in £s
In particular, she wishes to determine nonnegative prices per hour
Labour
of A-line 10 nonnegative
test capacity, 15 20 per
prices in hours
hour of 2000
C-line test
A-line
capacity 1
and nonnegative 1
prices 0
per hour inlabour
of hours capacity
120 to offer
Ajax that will induce
C-line 0 Ajax into
0 agreeing1 to rent its resources
in hours 48 rather
than to use them in the manufacture of its own products. At the
same time she wishes to pay the least amount for these resources. A
number of large manufacturing firms are concered with the economic viability of
their small suppliers.

3/10/2010 Tibor Illés – Optimisation - 2008 7

 Multiperiod Resource Allocation Problem
After reviewing the optimal assembly plan given by the optimal solution (120,
0, 40) of the previous LP model the Ajax’ marketing manager doubts that his
organization can sell 120 units of α-PCs next week. Moreover, he is concerned
that the plan does not call for the assembly of any β-NBs. Even if the profit
margin from β-NBs is smaller relative to the other two pruducts, the marketing
manager feels strongly that it must be included in Ajax’ product line.
Thus he requests instead that the production manager develop a 4-week
production strategy based on the sales forecasts shown
Because Ajax has capital tied up in Week 1 Week 2 Week 3 Week 4
products, carrying costs must be α-PC [20,60] [20,80] [20,120] [20,140]
charged for items held in inventory. β-NB [20,40] [20,40] [20,40] [20,40]
These carrying costs £4 per week for γ-WS [20,50] [20,40] [20,30] [20,70]
each α-PC, £5 per week for each β-NB, and £9 per week for each γ-WS. Initial inventory
at this week equals 22 α-PCs, 42 β-NBs, and 36 γ-WSs. For each product, these balance
equations are of the form
I(t) = I(t-1) + P(t) – S(t)
3/10/2010 Tibor Illés – Optimisation - 2008 8