Value Engineering and Operations

Optimization (AE4441)

Delft
University of
Technology

Challenge the future

Overview of the course

Period 1
Wednesdays and Thursdays: Lectures/studio classroom
sessions on optimization methods
Introduction to Operations Research by Frederick S.
Hillier and Gerald J. Lieberman, ninth edition
Finished with written exam in first period (week 1-10, not
yet scheduled). NOT open-book.

Period 2
Two assignments
Focus on value engineering approach
Solve a value engineering problem with optimization methods

Lectures on value engineering

Introduction

Overview of the course, continued

Final grade:
Exam holds for 60 % of the final grade
Each of the assignments holds for 20 % of the final grade
A grade > 5.5 for each of the elements is required to pass the
course
Grades are valid for one year only

Introduction

How to buy the book

The book can be bought via the VSV
Use needs to be made of the new VSV webshop. You can
find this webshop under the Book section at the website (
www.vsv.tudelft.nl). You can pay via iDeal. The book will
be delivered at the desired address.

Introduction

Overview of todays lecture

1st part
Chapters 1,2,3.1-3.4

2nd part
Assignments related to Chapter 3

Introduction

Chapter 1: Introduction
The field of Operations Research (OR) started in World War II
This research on operations was afterwards introduced in other
organizations
Operations Research aims at determining optimal ways to conduct
activities in an organization

Introduction

Chapter 2: Overview of the Operations

Research Modeling Approach
Typical phases in an OR study:
1. Defining the problem and gathering data
2. Construct a mathematical model to represent the problem

n decision variables: the unknowns

Objective function: measure of performance
Constraints: restrictions on decision variables
Parameters: constraint and objective function constants

3. Develop a computer-based procedure for deriving solutions

to the problem from the model
4. Test the model and refine if needed (model validation)
5. Prepare for ongoing application of the model (decision
support system)
6. Implement

Note: these steps also hold for other

fields. It holds for optimization approaches
in general.

Introduction

Chapter 2: Overview of the Operations

Research Modeling Approach
Typical phases in an OR study:
1. Defining the problem and gathering data
2. Construct a mathematical model to represent the problem

n decision variables: the unknowns

Objective function: measure of performance
Constraints: restrictions on decision variables
Parameters: constraint and objective function constants

3. Develop a computer-based procedure for deriving solutions

to the problem from the model
4. Test the model and refine if needed (model validation)
5. Prepare for ongoing application of the model (decision
support system)
6. Implement

Note: these steps also hold for other

fields. It holds for optimization approaches
in general.

Introduction

Example

Resource optimization for operations

and maintenance of offshore wind farms

1) Problem: Large-scale roll-out of offshore wind energy

requires that its cost of electricity should be reduced.
Potential for cost reduction in the field of operations and
maintenance.
2)The largest cost drivers are the vessels and technicians.
The following decision variables are identified:

Introduction

Resulting mathematical model:

Maximize
where

Y f ( A) f (C P ) f (C D )
A f ( x1 , x 2 , x3 , x 4 , x5 , x 6 )

the availability of the wind-farm

C P f ( x1 , x3 , x5 )
C D f ( x1 , x 2 , x3 , x 4 , x5 , x 6 ) days of preventive maintenance
costs (salaries and equipment)

such that A 95%

PM PM Re q

x1 1
x2 1
x3 1
x4 1

x5
x 6 minimum technician team size on a SOV
minimum technician team size on a CTV

, and

x1 , x 2 , x3 , x 4 , x5 , x6

are all integer values

Introduction

10

3) Result obtained by computer-based procedure for

deriving solutions to the problem from the model:

4) Testing on existing offshore windmill parks

5) Decision support tools for ongoing application (daily
operations and new offshore windmill farms)
6) Ensure use of the tools
Introduction

11

Linear Programming
Introduction

All functions in the model representing the problem are

linear

A method for efficiently solving linear programming

problems is available (1947, George Dantzig)
Note: the term programming does not refer to computer
programming but to planning

Chapter 3

12

Linear Programming
Introduction, continued

Most common application of linear programming:

How to allocate limited resources among competing activities in
a best possible way

Constraints

Decision variables

Objective
function

Chapter 3

13

An example problem
Define the problem

The WYNDOR GLASS CO. produces high-quality glass products, including

windows and glass doors. It has three plants.

Aluminum frames and hardware are made in Plant 1

Wood frames are made in Plant 2

Plant 3 produces the glass and assembles the products

Chapter 3

14

An example problem
Define the problem

Because of declining earnings, it has been decided to revamp the companys

product line. Unprofitable products are being discontinued. This releases
production capacity to launch two new products having large sales potential
Product 1: A glass door with aluminum framing
Product 2: A double-hung wood-framed window
Product 1 requires some of the production capacity in Plants 1 and 3, but none in
Plant 2. Product 2 needs only Plants 2 and 3
The marketing division has concluded that the company could sell as much of
either product as could be produced by these plants. However, because both
products would be competing for the same production capacity in Plant 3, it is not
clear which mix of the two products would be most profitable.

Chapter 3

15

An example problem
Gather data

First values for the parameters needed

Number of hours of production time available per week in each plant for the new products
Number of hours of production time used in each plant for producing the new products

Profit
for each of the new products

Production time per

batch (Hours)
Plant

Product

Production time
available per week
(Hours)

12

18

Profit per batch

$ 3000

$ 5000

Chapter 3

16

An example problem

Its linear programming model

Decision variables:

x1: number of batches of product 1 produced per week

x2: number of batches of product 2 produced per week

The objective function

Z = 3x1 + 5x2 (total profit per week in thousands of dollars)

The aim is to maximize Z

Constraints

x1 4 Number of hours of production time available per week in plant 1

2x2 12Number of hours of production time available per week in plant 2

3x1 + 2xNumber
of hours of production time available per week in plant 3
2 18

and

x1 0

x2 0

Chapter 3

17

An example problem

Its linear programming model

Maximize

Z = 3x1 + 5x2

Subject to
x1 4
2x2 12
3x1 + 2x2 18
x1 0
x2 0

Chapter 3

18

An example problem
Derive solutions

Maximize Z = 3x1 + 5x2

Subject
to
x1 4
2x2 12
3x1 + 2x2 18

Chapter 3

19

An example problem
Find optimal solution

Z = 3x1 +
5x2
Feasible
region

Z = 36
Z = 30
Z = 20

Chapter 3

20

An example problem
Optimal solution

For this situation a graphical approach suffices. The optimal solution is:
x1 = 2
x2 = 6
Z = 36
, i.e. maximal profit within the constraints is obtained for a mix of 2
batches of product 1 per week, and 6 batches of product 2 per week.
The resulting total profit is $36000 per week

Product 1: glass door with aluminum framin

Product 2: Double-hung wood-framed windo

Chapter 3

21

Linear Programming

The mathematical model, its general form

Maximize

Z = c1x1 + c2x2 + . + cnxn

Objective function

Subject to
a11x1 + a12x2 + + a1nxn b1
a21x1 + a22x2 + + a2nxn b2

Functional
constraints

am1x1 + am2x2 + + amnxn bm

and
x10, x20, , xn0

Nonnegativity
constraints

Note: other forms also possible

Chapter 3

22

Linear Programming

The mathematical model, its solutions

Feasible solution:
Nonfeasible solution:
is violated
Optimal solution:

All constraints are satisfied

At least one constraint

objective

Feasible solution with most

favourable value of the
function

Corner-point feasible solution:

(CPF)

Solution at the corner of the

feasible region

Chapter 3

23

Linear Programming

The mathematical model, its solutions

The importance of CPF solutions:

Any linear programming problem with feasible solutions
and a bounded feasible region must possess CPF
solutions and at least one optimal solution.
The best CPF solution must be an optimal solution.
If there are multiple optimal solutions, at least two must
be CPF solutions

Chapter 3

24

Linear Programming

The assumptions, proportionality

Proportionality:
The contribution of each activity j to the value of the
objective function Z is proportional to the level of activity
xj (cjxj)
and
The contribution of activity j to the left-hand-side of each
functional constraint is proportional to the level of
activity xj (aijxj) Proportionality violated

These situations
prevent a
(straightforward) linear
programming approach
to solve for optimal
solutions for the
decision variables.

Chapter 3

25

Linear Programming
The assumptions, additivity

Additivity:
Every function in a linear programming model is the sum
of individual contributions of the respective activities
Functions that do not satisfy this requirement, e.g.:
Z = 3x1 + 5x2 - x1x2
Z = 3x1 + 5x2 + x1x2

Chapter 3

26

Linear Programming

The assumptions, divisibility

Divisibility:
Decision variables in a linear programming model are
allowed to have any values, including non-integer values,
that satisfy the functional and non-negativity constraints

Chapter 3

27

Summary
Problems dealing with assigning scarce resources to a
series of activities in an optimal way, can often be solved
through linear programming

Hereto a mathematical model is established

The objective function reflects the measure of
performance
The constraints reflect the limited resources
A search needs to be carried to find those levels of
activities that maximize the objective function and still
are within the constraints
Chapter 3 considers a graphical approach. This is not
feasible when more than three levels of activity need to
be determined
Chapter 3 28

Summary, continued
The feasible region is the collection of all possible
solutions (levels of activity) that satisfy all constraints
Corner-point feasible (CPF) solutions are located at a
corner of the feasible region
These CPF solutions are of interest because the best
CPF solution must be an optimal solution. If the problem
has multiple optimal solutions at least two must be a CPF
solution
The assumption of a linear objective function subject to
linear constraints is often allowed for real life problems
Chapter 3

29