Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

General info:

Questions either in class or in the forum


No need for a textbook
-> lots of exercises to practice
Chapters covered 1 and 3-7 ( not everything from them)
HOMEWORK
Send by E-MAIL -> Important for grading

Exam structure ( 5 questions)


1) Formulate an LP model for a given problem description
2) Solve a given LP model using the simplex method
3) Formulate the dual of a given LP model
4) Apply sensitivity analysis to a given LP model and solution
5) Mystery -> hard one

It is possible that software questions would be asked on the exam, to interpret what is
going on there

https://p.cygnus.cc.kuleuven.be/bbcswebdav/pid-33015417-dt-content-rid-333154622_2
/courses/B-KUL-HBE07e-2223/LO0%20-%20Practical%20stuff.pdf -> more info here
Lecture 1
How to convert a story into a Linear optimization problem:
1) Example
a) You have 6 big blocks and 8 small ones. You can construct a table and a chair. A
table requires you to use 2 big and 2 small blocks, whereas a chair requires you
to use 1 big and 2 small blocks. If a chair could be sold for 100$ and a table for
160$, maximize the profit.
2) Method
a) First, describe a possible solution
i) How many tables/chairs are we going to construct
(1) X(t) -> table / X(c) -> chairs
ii) What is the thing you want to maximize
(1) Profit
(2) P(x,y) = 160xT + 100 xC
(a) Here you compare solutions
iii) Then determine the constrain
(1) 2xT + 2xC <= 8 ( table/chair require 2 small blocks to be build)
(2) 2xT + xC <= 6
(3) xT >=0
(4) xC => 0
(5) Very important the red part
(6) Also, it is always bigger/smaller OR equal
(a) Strictly smaller (<) or bigger (>) are
not linear !!! (negative points)
b) End result
Terminology
1) decision variable -> X(t) and X(c)
2) objective (or goal) function -> maximize 160xT + 100xC
3) Constraint -> Subject to part
4) non-negativity constraint -> x(t) >= 0
5) unrestricted in sign -> No info is given on the sign
6) technological coefficient -> 2xT + 2xC <= 8
7) goal function coefficient -> 160xT + 100 xC
8) SHADOW PRICE -> value that an extra block creates
a) For example, here 1 extra big block allows us to build 1 more table
i) So 160 - 100 = 60$ is the shadow price of that extra big block
b) A small block doesn’t help us at all, so the shadow price is 0

When do we have a non-linear model?


1) When we multiply/divide the decision variables with one another
a) X1/ x2 <= 10
2) When we have sin, cos, x^2, Sqrt(x)
3) STRICT INEQUALITIES (x<5 or y >3,5 are NOT LINEAR)
Lecture 2 - Good tips
1) Use a multi-dimensional index when variables have a natural organization in rows &
columns; or links between sets of objects/agents
2) Example:
a) given 3 workers (Tom, Ben, Matt) and 2 machines (1, 2), represent the number of
minutes each worker operates each machine as:
b) xT1, xT2, xB1, xB2, xM1, and xM2

Types of problems
• A diet problem (3.4)
• A work scheduling problem (3.5)
• A capital budgeting problem (3.6)
• Short-term financial planning (3.7)
• Blending problems (3.8)
• Production process models (3.9)
• A cutting problem (-)
• Inventory problems (3.10)
• Multi-period financial model (3.11)
• Multi-period work scheduling problem (3.12)
• Transportation problem (7.1)
• Assignment problem (7.5)
• Transshipment problem (7.6)

A Work-Scheduling Problem - how to think of it


The key to correctly formulating this problem is to realize that the post office’s primary
decision is not how many people are working each day but rather how many people begin work
each day of the week.
Using Linear Programming to Solve Multiperiod Decision
Problems: An Inventory Model
Decisions made during the current period influence decisions made during future
periods. In formulating any multiperiod LP model, the hardest step is usually finding the
relation that relates decision variables from different periods.
Step 1:
Determine the number of sailboats that should be produced (both regular and irregular)
x(t) (t=1,2,3,4) -> inventory produced in period “t” with a regular price (Ex: 400)
y(t) (t = 1,2,3,4)-> inventory produced in period “t” with a overdue price (Ex: 450)

Step 2:
Determine a decision variable that shows the inventory at the end of each period
E(t) (t=1,2,3,4) -> #of inventory at the end of period “t”

Step 3:
Construct the total cost function:
Total cost = (cost of producing regular-time invent) + cost of producing overtime invent.
+ inventory costs

Step 4: ( IMPORTANT OBSERVATIONS )


Inventory at end of quarter t ==
inventory at end of the quarter (t - 1) + quarter (t) production - quarter (t) demand
E(t) = E(t-1) + (x(t)+ y(t)) - D(t)

Since demand has to be met, we need E(t) >=0,

Step 5: Constraints
x1,x2,x3,x4 <= 40 ( cap on regular production)
E(1) = 10 +x1 + y1 - 40 # the 10 comes from the inventory at the start of Q(1)
E(2) = E(1)+ x2 + y2 - 60
E(3) = E(2) + x3 + y3 - 75
E(4) = E(3) + x4 + y4 - 25
E(t) >= 0 # so that we meet demand
x(t), y(t) >= 0 # because we cannot unproduce

To be certain that the demand for the period is met, the left-over inventory has to be
more than or equal to 0
In (t) >0 for t = (1,2,3,4)
Transportation problem (7.1)

4 Calculating the total cost is usually distance traveled times units transported. This
is usually the function we would try to minimize.

5 Constraints
1) The total shipments from the supply points CAN NOT exceed their capacity
a) X(1,1) + x(1,2) +x(1,3)+ x(1,4) <= Capacity (1)
b) X(2,1) + x(2,2) +x(2,3)+ x(2,4) <= Capacity (2)
2) Each demand point ( city) will receive at least the needed amount
a) X(1,1) + X(2,1) + X(3,1) >= Demand (1)
b) X(1,2) + X(2,2) + X(3,2) >= Demand (2)
3) The units supplied should be NON-NEGATIVE
a) X(i,j) >=0
3.11 Multiperiod Financial Models
The key is to determine the relations of cash on hand during different periods that the
Finco Investment Corporation must determine an
investment strategy for the firm during the next three
years. Currently (time 0), $100,000 is available for
investment. Investments A, B, C, D, and E are available.
The cash flow associated with investing $1 in each
investment is given in Table 38.
On the other hand, keeping money at hand earns an
8% interest rate.

1) Each investment at first requires a $1 Dollar


outflow
2) Some investments produce inflow the next time,
while others will (not seen here)
a) Then at a later stage, the cash produced
by the investment can be immediately reinvested in the next period to receive
further cash.
3) Don’t forget the cash in mutual markets that earns 8% interest

Decision variables
1) A dollar invested in investment A (A <=75000
2) B dollars invested in investment B (B<=75000
3) C dollars invested in investment C (C <=75000
4) D dollars invested in investment D (D <=75000
5) E dollars invested in investment E (E <=75000
6) St dollars invested in money market funds at time t (t 0, 1, 2)

We want to maximize cash at hand at time 3


Max Z = 1B + 1.9D + 1.5E + S2 *10.8

Subject to
1) 100,000 = A + C + D + S0 -> We can invest 100 000 in those investments and the
left-over is just saved as cash
2) 0.5A + 1.2C +1.08S0 = B + S1 ( The cash inflow from the investment can be spread in
the new opportunities)
3) A + 0.5B + 1.08S1 = E + S2
4) And we have that none of them should be more than 75000
5) And all should be NON-NEGATIVE !!!!

You might also like