Programming Class 1

Model-Building Tools

Model-Building Tools

Escola Superior Náutica Infante D. Henrique – ENIDH

2023/24

Camilo Lima Programming

 Model-Building Tools Contents

Contents

• Solving linear programming (LP) problems using the Excel solver.

Camilo Lima Programming

 Model-Building Tools
Introduction

A Quick Introduction

• Managerial issues: the key to profitable operations is making the best use of available resources
of people, material, plant & equipment, and money.
• Technique: a mathematical modeling tool is available for this purpose with linear programming (LP).
• Course objective: to show how the use of the Excel Solver to solve LP problems opens a whole
world to managers and provides an invaluable addition to his/her technical skill set.
• Course organization: a quick introduction to LP is started. Then, LP problems are later
introduced.

Camilo Lima Programming 3

 Model-Building Tools
Introduction

Linear Programming (LP)

• Decision-making process: managers often must make decisions on how best to allocate scarce
resources among competing activities.
• Highlight: scarce resources (e.g., machinery, time, etc.) are used to manufacture various products or
provide diverse services.
• The management’s problem: to find the best way to produce products, given the scarcity of resources.

• Linear programming (LP): a mathematical technique designed to aid in the planning and
allocation of key organizational resources.
• Linear: all equations and inequalities associated with the problem have linear relationships.
• Programming: the iterative process that is used to derive an optimum solution.

Camilo Lima Programming 4

 Model-Building Tools
Introduction

Linear Programming (LP)

• Highlight: all LP problems are divided into two parts.

• An objective: LP problems aim at maximizing or minimizing some quantity, such as profit or cost.
• Example: how to determine the optimal allocation of funds in an investment portfolio to maximize profits.
• A set of constraints: the resource limitations, called constraints, are described in terms of linear
inequalities, expressed either as ≥ (greater-than-or-equal-to) or ≤ (less-than-or-equal-to).
• Corollary: the presence of resource inequalities creates alternative solutions to the LP problem, with (usually) only
one solution being optimal.

Camilo Lima Programming 5

 Model-Building Tools
Introduction

Linear Programming (LP)

• The best-known linear programming methods: the Simplex method and the graphical method.
• The graphical method: used only for problems with two or three variables. Therefore, it is of limited use.
• The simplex method: it can solve very large LP problems involving thousands of variables. It consists of a
set of step-by-step rules, which are ideally suited for computerization.

Camilo Lima Programming 6

 Model-Building Tools
Introduction

The Linear Programming Model

• The linear programming problem entails an optimizing process in which non-negative

values for a set of decision variables X1, X2, . . . , Xn are selected to maximize (or minimize) an
objective function in the form:

Maximize Z = C1 X1 + C2 X2 + . . . + Cn Xn
Subject to resource constraints in the form:
A11 X1 + A12 X2 + ... + A1n Xn ≤ B1
A21 X1 + A22 X2 + ... + A2n Xn ≤ B2
.
.
.
Am1 X1 + Am2 X2 + ... + Amn Xn ≤ Bm

• Where Cn, Amn and Bm are given constants. Depending on the problem, the constraints also may
be stated with equal signs (=) or greater-than-or-equal-to (≥) signs (in a minimize direction).

Camilo Lima Programming 7

 Model-Building Tools
Introduction

The Linear Programming Model

• The five requirements for standard LP:

• (1) Limited resources (e.g., a finite number of hours available at each machine center).
• (2) Explicit objective function (e.g., maximize profit or minimize cost).
• (3) Linearity (i.e., there are no exponents or cross-products).
• (4) Homogeneity (i.e., everything is in one unit of measure).
• (5) Divisibility (i.e., decision variables are divisible and nonnegative).

Camilo Lima Programming 8

 Model-Building Tools
Introduction

Applications of Linear Programming

• Linear programming problems fall into two main categories:

• Blending or product-mix problems: they are focused on achieving an optimum combination of
resources to maximize profits or minimize costs.
• Example: a typical food-blending problem could involve a developing country wishing to produce a high-protein
food mixture at the lowest possible cost.
• Transportation problems: they involve supply sources (warehouses) and destinations (shops). The
objective is to minimize the transportation costs to move wholesale products from warehouses to shops.
• Example: the transportation problem that concerns selecting minimum-cost routes in a product-distribution network
between sources and destinations.

Camilo Lima Programming 9

 Model-Building Tools
Introduction

Applications of Integer Programming and Goal Programming

• Integer numbers: often, solutions to LP problems make sense only if they have integer values.
• Example: quantities such as 23!⁄" employees or 12!⁄" tables are unrealistic. Simply rounding off the LP
solution to the nearest whole number may not produce a feasible solution.
• Integer programming (IP): the term used to describe an LP problem that requires the solution to have
integer values.
• Multiple objectives: a limitation of LP is that it allows only one objective function to be optimized.
Thus, the decision-maker must focus on a single objective or goal at a time.
• Example: how to maximize both production and profits or to maximize profits while minimizing risk.
• Goal programming (GP): a variation of LP that provides for multiple objective optimization.

Camilo Lima Programming 10

 Model-Building Tools
Introduction

Applications of Nonlinear Programming (NLP)

• Nonlinear programming (NLP): a technique capable of solving problems containing nonlinear

relationships in their equations and/or inequalities.
• (1) A nonlinear relationship is any equation that does not conform to the straight-line relationship, i.e., y =
mx + c.
• (2) Formulating an NLP problem is the same as formulating an LP problem, but the objective and/or
constraints may be nonlinear.

Camilo Lima Programming 11

 Model-Building Tools
Introduction

Excel’s Solver

• Excel’s Solver: most spreadsheet packages have built-in solvers.

• (1) A solver is a powerful tool for analyzing and solving various types of LP problems.
• (2) Excel’s solver can solve three types of problems, i.e., LP, IP, and NLP.
• (3) Due to its iterative nature, LP may not always converge to the best solution.
• (4) Given the complexity of NLP, Solver cannot ensure the optimality of the solution it identifies.

Camilo Lima Programming 12

 Model-Building Tools

References

Suggested Books

• Barlow, J.F., 2005. Excel Models for Business and Operations Management, Second Edition. John
Wiley & Sons Ltd.
• Jacobs, F.R., Chase, R.B., 2018. Operations and Supply Chain Management, Fifteenth Edition.
McGraw-Hill Education.
• Powell, S.G., Baker, K.R., 2014. Management Science: The Art of Modeling with Spreadsheets
Fourth edition. John Wiley & Sons.
• Ragsdale, C., 2008. Spreadsheet Modeling & Decision Analysis: A Practical Introduction to
Management Science, Fifth Edition. Thomson South-Western.

Camilo Lima Programming