Topic 7 (Ch8)

Nonlinear Programming &

Evolutionary Optimization
 Introduction to Nonlinear Programming
(NLP)
 An NLP problem has a nonlinear objective
function and/or one or more nonlinear
constraints.

 NLP problems are formulated and implemented in

virtually the same way as linear problems.
Possible Optimal Solutions to NLPs
(not occurring at corner points) Examples of NLPs
objective function objective function
level curve level curve

optimal solution optimal solution

Feasible Feasible
Region Region

linear objective, nonlinear objective,

nonlinear constraints linear constraints

objective function
level curve objective function
level curves
optimal solution
optimal solution

Feasible
Region
Feasible
Region
nonlinear objective, nonlinear objective,
nonlinear constraints linear constraints
 An NLP Solution Strategy

X2 C D
B E

objective function
level curves

Feasible
Region
A
(the starting point)

X1
Local vs. Global Optimal Solutions

X2
Local optimal solution

C
Local and global
F optimal solution
E
Feasible Region
G
B

A
D

X1
1. The Economic Order Quantity
(EOQ) Problem
 A common business problem
 Determining the optimal quantity to purchase
when orders are placed.
 EOQ problem assumptions:
– Demand (or use) is constant over the year.
– New orders are received in full when the
inventory level drops to zero.
 The EOQ Problem
 Large orders (fewer purchase orders) result in:
– higher inventory levels & carrying costs
– infrequent orders & lower ordering costs
 Small orders (more purchase orders) result in:
– low inventory levels & carrying costs
– frequent orders & higher ordering costs

7
 Sample Inventory Profiles
Inventory
60
Annual Usage = 150 Number of Orders = 3
50 Order Size = 50 Avg Inventory = 25

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 Month
Inventory
60
Annual Usage = 150 Number of Orders = 6
50 Order Size = 25 Avg Inventory = 12.5

40

30

20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 Month
 The EOQ Model
Three kinds of costs:
1. Payment for goods Total: 1+2+3
2. Ordering cost
3. Carrying cost
D Q
Total Annual Cost = DC  S  Ci
Q 2
where:
D = annual demand for the item
C = unit purchase cost for the item
S = fixed cost of placing an order
i = cost of holding inventory for a year (expressed as a
% of C)
Q = order quantity
 EOQ Cost Relationships
$

1000

800 Total Cost

600

400 Carrying Cost

200 Ordering Cost

EOQ
0
0 10 20 30 40 50
Order Quantity
 An EOQ Example:
Ordering Paper For MetroBank
 Alan Wang purchases paper for copy machines
and laser printers at MetroBank.
– Annual demand (D) is for 24,000 boxes
– Each box costs $35 (C)
– Each order costs $50 (S)
– Inventory carrying costs are 18% (i)
 What is the optimal order quantity (Q)?
 The Model
D Q
MIN: DC  S  Ci
Q 2
Subject to: Q 1
(Note the nonlinear
objective!)
Implementing the Model

See file Fig8-6.xlsm

 2. Location Problems
 Many decision problems involve determining optimal
locations for facilities or service centers. For example,
– Manufacturing plants
– Warehouse
– Fire stations
– Ambulance centers
 These problems usually involve distance measures in the
objective and/or constraints.

 The straight line (Euclidean) distance between

two points (X1, Y1) and (X2, Y2) is:

Distance  X 1  X 2   Y1  Y2 
2 2
 A Location Problem:
Rappaport Communications
 The company is planning to offer cellular
service to several cities.
 They want to expand to provide inter-city
service between four cities in northern Ohio.
 A new communications tower must be built to
handle these inter-city calls.
 The tower needs to be located within 40 miles
of each of these cities. (40 miles - the limit of
the broadcast radius of the new equipment)
Graph of the Tower Location Problem
Y
50
Cleveland
x=5, y=45
40

30
Youngstown
Akron
x=12, y=21 x=52, y=21
20

10
Canton
x=17, y=5
0
20 30 40 50 60
X
0 10
Defining the Decision Variables

X1 = location of the new tower with

respect to the X-axis
Y1 = location of the new tower with
respect to the Y-axis
 Defining the Objective Function
 Minimize the total distance from the new
tower to the existing towers

       
5-X
1
2
 45  Y
1
2
 12 - X
1
2
 21  Y
1
2

MIN:
       
 17 - X
1
2
 5 Y
1
2
 52 - X
1
2
 21  Y
1
2
Defining the Constraints
 Cleveland
5 - X12  45  Y1 2 40
 Akron
 12 - X
1
 
2
 21  Y
1
 2
 40
 Canton


17 - X
 Youngstown
1
  
2
 5 Y
1
2
 40

   
52 - X
1
2
 21  Y
1
2
 40
Implementing the Model
See file Fig8-10.xlsm
 Analyzing the Solution
 The optimal location of the “new tower” is in the
same location as the existing Akron tower.
 Maybe they should just upgrade the Akron
tower.
 The maximum distance is 39.8 miles to
Youngstown (near 40 mile transmission radius).
 Better to have a safety distance to ensure
quality and reliability during inclement weather.
 Where should we locate the new tower if we
want to minimize the maximum distance from
the new tower to each of the existing towers?
Defining the Constraints
 Cleveland
5 - X12  45  Y1 2 40
 Akron
 12 - X
1
 
2
 21  Y
1
 2
 40
 Canton


17 - X
 Youngstown
1
  
2
 5 Y
1
2
 40

   
52 - X
1
2
 21  Y
1
2
 40

22
 Objective function
 Min: D
 Cleveland
5 - X12  45  Y1 2 40D
 Akron

12 - X
1
  2
 21  Y
1
 2

D
 40
 Canton


 
17 - X
Youngstown
  1
2
 5 Y
1
2

D
 40

   
52 - X
1
2
 21  Y
1
2

D
 40

23
Implementing the Model
Revise the file Fig8-10.xlsm
COMPARISON OF THE TWO SOLUTIONS

This solution positions the new tower at the X-Y coordinates

(26.84, 29.75). Although the total distance associated with this
solution increased to 97.137 (or an average of 24.28 miles),
the maximum distance was reduced to about 26.6 miles. 25
3. A Nonlinear Network Flow Problem:
The SafetyTrans Company
 SafetyTrans specialized in trucking extremely
valuable and extremely hazardous materials.
 Not interested in routes for the least time or cost, but
likes the routes that have a low risk in accidents.
 The company maintains a database of highway
accident data which it uses to determine safest
routes.
 They currently need to determine the safest route
between Los Angeles and Amarillo.
 Network for the SafetyTrans Problem
Las
Vegas 0.006 0.001 Albu-
2
querque 0.001 +1
Flagstaff 8
6 Amarillo
10
0.003 0.010
0.006
0.004

Los
Angeles 0.002 0.009 0.010
1 Phoenix 0.005 0.006
4
-1
0.004 0.002
0.002 Lubbock
0.003 9
Las
San 0.003 Cruces
Diego Tucson 7
3 5
0.010

Numbers on arcs represent the probability of an accident

occurring.
Defining the Decision Variables

1, if the route from node i to node j is selected

Yij 
0, otherwise
 Defining the Objective
Select the safest route by maximizing the
probability of not having an accident,

MAX: (1-P12Y12)(1-P13Y13)(1-P14Y14)(1-P24Y24)…(1-
P9,10Y9,10)
where:
Pij = probability of having an accident while
traveling between node i and node j
 Defining the Constraints
 Flow Constraints
-Y12 -Y13 -Y14 = -1 } node 1
+Y12 -Y24 -Y26 = 0 } node 2
+Y13 -Y34 -Y35 = 0 } node 3
+Y14 +Y24 +Y34 -Y45 -Y46 -Y48 = 0 } node 4
+Y35 +Y45 -Y57 = 0 } node 5
+Y26 +Y46 -Y67 -Y68 = 0 } node 6
+Y57 +Y67 -Y78 -Y79 -Y7,10 = 0 } node 7
+Y48 +Y68 +Y78 -Y8,10 = 0 } node 8
+Y79 -Y9,10 = 0 } node 9
+Y7,10 +Y8,10 +Y9,10 = 1 } node 10
Implementing the Model
See file Fig8-15.xlsm
Comments on Nonlinear Network Flow Problems

 Small differences in probabilities can mean large differences in

expected values:
0.9900 * $30,000,000 = $300,000
0.9626 * $30,000,000 = $1,122,000
 4. A Project Selection Problem:
The TMC Corporation
 TMC needs to allocate $1.7 million of R&D budget
and up to 25 engineers among 6 projects.
 The probability of success for each project depends
on the number of engineers assigned (Xi) and is
defined as:
Pi = Xi/(Xi + ei)
Project 1 2 3 4 5 6
Startup Costs $325 $200 $490 $125 $710 $240
NPV if successful$750$120 $900 $400$1,110$800
Probability
Parameter ei 3.1 2.5 4.5 5.6 8.2 8.5
(all monetary values are in $1,000s)
 Selected Probability Functions
Prob. of Success
1.0000

0.9000 Project 2 - e = 2.5

0.8000
Project 4 - e = 5.6
0.7000

0.6000

0.5000 Project 6 - e = 8.5

0.4000

0.3000

0.2000

0.1000

0.0000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Engineers Assigned
 Defining the Decision Variables
1,if project i is selected
Yi  i 1, 2, 3, ..., 6
0,otherwise

Xi = the number of engineers assigned to project i, i = 1, 2,

3, …, 6
 Defining the Objective
Maximize the expected total
NPV of selected projects

750X1 120X 2 900X 3 800X 6

MAX:   
(X1  31
. ) (X 2  2.5) (X 3  4.5) (X 6  8.5)
 Defining the Constraints
 Startup Funds
325Y1 + 200Y2 + 490Y3 + 125Y4 + 710Y5 + 240Y6 <=1700
 Engineers
X1 + X2 + X3 + X4 + X5 + X6 <= 25
 Linking Constraints
Xi - 25Yi <= 0, i= 1, 2, 3, … 6

 Note: The following constraint could be used in

place of the last two constraints...
X1Y1 + X2Y2+ X3Y3+ X4Y4+ X5Y5 + X6Y6 <= 25
However, this constraint is nonlinear. It is
generally better to keep things linear where
possible.
Implementing the Model
See file Fig8-19.xlsm
Follow-up question #1

39
 Follow-up question #2
 Suppose a gift shop in Myrtle Beach has an annual
demand for 15,000 units for a souvenir kitchen magnet
that it buys for $0.50 per unit. Assume it costs $10 to
place an order and the inventory carrying cost is 25% of
the item’s unit cost. Use Solver to determine the optimal
order quantity if the company wants to minimize the total
cost of procuring this item.
 a. What is the optimal order quantity?
 b. What is the total cost associated with this order
quantity?
 c. What are the annual order and annual inventory
holding costs for this solution?
40
Follow-up question #3

41