Deterministic Models in Operations
Research
¨Instructor: Dr. Nguyen Van Hop
1
Chapter 1: Introduction to Operations
Research
¨Introduction
¨The Model
¨Procedures in OR
¨Develop the solution for decision making
1. Introduction
What is operation research?
¨ Operations Research (OR)
¨ OR searches for
optimal or best solutions
in the presence of resources constraints.
¨ OR uses mathematical techniques to model and
analyze issues related to decision making processes.
4
ART AND SCIENCE
In OR:
¨Science: OR provides mathematical techniques,
algorithms, and procedures to solve problems.
¨Art: before and after the searching for solution
are the creativity, reasoning ability, and personal
ability of the person who analyze and build the
model for the problem.
5
2. THE MODEL
¨Models are brief reflections of the real world.
¨Types of models:
¨ Mathematical model,
¨ Physical model,
¨ Simulation model
¨…
6
Question
¨What items are the models?
A. A map
B. A House
C. A Design Drawing
D. A car
E. A chemical equation: CH4+ O2 CO2 + H2O
3. Problem Solving Process
Goal: solve a problem Formulate the
Problem
Situation Problem
Model must be valid Implement a Solution Statement
Model must be tractable
Solutions must be useful Data
Construct
a Model
Implement
the Solution
Model
Solution
Find
a Solution
Establish
a Procedure
Test the Model
and the Solution Solution Tools
8
Procedure in Operations Research
¨Give an example how to make a cup of coffee?
¨What are key steps ?
9
Problem Definition
¨ Define the decision maker’s problem: clearly,
correctly
¨ Advisory capacity, appropriate objectives: who
make the decisions, long-run, satisfactory
profits..
¨ Look through the symptoms to find exactly the
“true” problem
¨ Focus to solve few problems
10
Formulate a Mathematical Model
¨ Types of Model: physical model, pilot, schematic, mathematical
model
¨ Define variables
¨ Define constraints
¨ Objective function: tractable model.
¨ Data requirements: data collection Complete, accurate,Reliable
¨ Deterministic data: solid, unchanged over time deterministic
models.
¨ Stochastic/probabilistic data: change randomly probabilistic
models.
¨ Example: Inventory model demand rate, lead time
11
Example
¨ Suppose your have 3 routes to go from your house to IU, How do you
plan for that job ?
¨ xi ={0,1} = 1 if I choose route I to go, 0; i =1,2,3
¨ Objective function: minimize time z=ax1+bx2+cx3
¨ Constraints:
¨ HR: No
¨ Physical item: no
¨ Facility: yes. Bus, motocycle, taxi
¨ Time: a,b,c: ax1+bX2+cx3<= 30 min
¨ Money: Ba,Bb, Bc
Bax1+BbX2+BcX3 <=20k
¨ Information: relationship choose 1 route: x1+x2+x3 =1
¨ X1=1, X2 =0, x3=0
12
Model Developing
¨ Decision variables
¨ Objective function
¨ Constraints
13
An Example of an LP model
¨ ABC company manufactures two products: Item A and Item B :x1, x2
Total profit =(revenue of product A – RM cost of A – labor cost of A) + (revenue of product B – RM cost
B –labor B) = (27x1 – 19x1 – 5x1) + (21x2 – 9x2 – 10x2) = 3x1 + 2x2
Finishing hour 2x1 + 1 x2 <= 100
Carpentry hour 1x1 + 1x2 <= 80
x2>=0
0<=x1<=40
¨ Each Item A:
¨ Sell for $27 and uses $19 worth of raw materials.
¨ Increase ABC’s variable labor/overhead costs by $5.
¨ Requires 2 hours of finishing labor.
¨ Requires 1 hour of carpentry labor.
¨ Each Item B:
¨ Sell for $21 and used $9 worth of raw materials.
¨ Increases ABC’s variable labor/overhead costs by $10.
¨ Requires 1 hour of finishing labor.
¨ Requires 1 hour of carpentry labor.
¨ Each week ABC can obtain:
¨
ABC’s LP model
¨ Variables:
¨ x1 = number of item A produced each week
¨ x2 = number of item B produced each week
¨ Objective function: Max z = 3x1 + 2x2 (1)
(Weekly profit = weekly revenue – weekly raw material costs – the weekly variable costs)
¨ Constraints: Subject to (s.t.)
¨ Each week, no more than 100 hours of finishing time may be used.
2 x1 + x2 ≤ 100 (finishing constraint) (2)
¨ Each week, no more than 80 hours of carpentry time may be used.
x1 + x2 ≤ 80 (carpentry constraint) (3)
¨ Because of limited demand, at most 40 item A should be produced.
x1 ≤ 40 (constraint on demand for Item A) (4)
x1 ≥0 (sign restriction) (5)
x2 ≥ 0 (sign restriction) (6)
MATHEMATICAL MODEL
¨ Models are categorized by:
¨ Single vs. multi-objective function
¨ Constraints: linear or non-linear
¨ Variables:
¨ Deterministic vs. Probabilistic (Stochastic)
¨ Discrete vs. Continuous
¨ Others: fuzzy, fuzzy stochastic…
¨ Parameters:
¨ Deterministic vs. Probabilistic (Stochastic)
¨ Discrete vs. Continuous
¨ Others: fuzzy, fuzzy stochastic…
Example:
¨ Linear programming. - Integer programming
¨ Non-linear programming - Markov chains 16
Solution Development
Solution:
¨ Multi-equation solving
¨ Trial and errors
¨ Enumeration
¨ Algorithm
Trial solutions:
¨ before and after apply solution
¨ precision and tractability of model
¨ robustness
¨ model validation
17
Results Analysis
Result Analysis :
¨ Sensitivity analysis
¨ postoptimality analysis: what-if analysis
Test the Model:
¨ model validation
¨ retrospective test
Implementation
¨ Simulation necessary
¨ Phase-by-phase: careful explanation, obtain feedback
¨ Experiments: clear and accurate methodology : reproducible ,
replicability
“Wrong” application leads to terrible results
18
4. DEVELOPING SOLUTIONS & MAKING
DECISIONS
Decision making is tough, challenging, but interesting.
Decision making includes:
¨Quantitative approach
¨Qualitative approach
¨Objectives of organizations
¨Constraints
¨Priority rules
19
Example 1
Select the optimal solution based on some criteria – available
alternatives
A senior manager of an organization wants to recruit a fresh
graduate to the position of demand planner. There are 3
candidates A, B, and C.
Question:
Give three main criteria to select the best candidate?
20 20
Example 2
Find the optimal combination of products to earn highest profit.
A company which produces office furniture whose main
products are tables and chairs. The production process
requires specific working hours for carpentry and paint. Data
of profit, required working hours of each product are
available.
Question:
What is the optimal combination of products for the company to
earn highest profit?
21 21
