Chapter 5

Network Models
What to be addressed:
 Special LP models: Optimization models with a
graphical network representation
2 5

1 7
4
8
3
6

Node: generally a location

Arc: generally a route 2
Areas of applications:

 Transportation
 Logistics
 Others …

3
Transportation Model:
Network representation
Origins Destinations

5
2

6
3
7 4
Transportation Models
 Companies produce products at plant locations (called
origins) and ships these products to customer locations
(called destinations).
 Each origin has a limited amount that it can ship, and
each customer destination must receive a required
quantity of the product.
 Products are directly shipped from an origin to a
destination.
 Different costs incur for different ways of shipping.

5
Transportation Model:
Network representation
Origins Destinations

5
2

6
3
7 6
Transportation Problems
 A typical transportation problem requires three sets of numbers:
– Capacities (or supplies) – indicates the most each plant can supply in a
given amount of time.

– Demands ( or requirements

– Unit shipping (and possibly production) costs

7
Example 5.1 (a) – Transportation Problem
 Grand Prix manufactures automobiles in three plants and then
ships them to four regions of the country.
 The production capacity of each plant, the demand in each
region, and the unit shipping cost between each plant/region
are given below:

Input data for Grand Prix Example

Region 1 Region 2 Region 3 Region 4 Capacity

Plant 1 131 218 266 120 450

Plant 2 250 116 263 278 600
Plant 3 178 132 122 180 500
Demand 450 200 300 300

8
 Example 5.1(a) – Transportation
Problem
 Grand Prix wants to find the lowest-cost shipping plan for
meeting the demands of the four regions without
exceeding capacities of the plants.

9
Network representation

10
Formulation
 Decision variable: The number of automobiles shipped from
each plant to each Region
 Requirement:
– The total number shipped out of each plant <= plant capacity

– The total number received by each region >= region demand

 Objective: The total shipping cost incurred

11
Solution

12
Supply Less Than Demand
 If total capacity is less than total demand, then there is no way
to meet total demand – Solver will report “no feasible solution.”
 We need to change the model.
 We need to drop the “>= demand” constraints and probably
include unit penalty costs for not meeting demand at the various
cities.

13
Example 5.1(b) – Transportation Problem
 Suppose the total supply capacity is less than the total
demand.
 The production capacity of each plant, the demand in each
region, the unit shipping cost between each plant/region, and
the unit shortage penalty cost in each region are given below:
Input data for Grand Prix Example

Region 1 Region 2 Region 3 Region 4 Capacity

Plant 1 131 218 266 120 450

Plant 2 250 116 263 278 350
Plant 3 178 132 122 180 300
Demand 450 200 300 300

Shortage
Cost 500 700 900 850

 Revise the previous model to find an optimal shipping plan.

14
“Disallowed” Routes
 Suppose that, for whatever reason, flows from certain plants
to certain cities are not allowed.
 How would you take of this in the formulation.

15
Exercise: Production and Shipment with Tax Consideration
 Suppose the Grand Prix’s problem is extended as follows: (1)
unit production costs in different plants are different; (2) unit
selling prices in different regions are different; (3) there are
different tax rates for different plants on their profits, as given
below:

Input data for Grand Prix Example

Region 1 Region 2 Region 3 Region 4 Capacity Unit Prod. Cost Tax Rate

Plant 1 131 218 266 120 450 14,450 30%

Plant 2 250 116 263 278 600 16,270 35%
Plant 3 178 132 122 180 500 16940 22%
Demand 450 200 300 300
Selling price $19,290 $20,520 $17,570 $18,320

16
Find the optimal production and shipping plan for Grand Prix.
Ex. 5.1(cont’d) – Alternative Model
 List all of the arcs and their corresponding flows in one
long list.
 Have one changing cell per arc.

 Need to use SUMIF() function

17
Ex. 5.1(cont’d) : An Alternative Model
 Some standard terminology.
 When we represent a network model graphically,
we generally connect node with arrows.
 Flow (in an arc): amount shipped through an arc

– Inflow: flow in arcs pointed into a node

– Outflow: flow in arcs pointed out of a node

 Total amount shipped out of an origin

= Total out flow of the origin

 Total amount received by a destination

= Total inflow of the destination

18
Ex. 5.1(cont’d) : An Alternative Model
 Key Excel function: to compute the total inflow of destination
and total outflow of origin

– SUMIF(Range,Criteria,SumRange)

19
Solution

20
Ex. 5.1(cont’d) : An Alternative Model
 An additional benefit from this model formulation is that it
makes it easy to disallow routes.
 This is very valuable when the number of potential arcs in the
network is huge – even though the vast majority of them are
disallowed – and this is exactly the situation in most large
network models.

21
Assignment Models
 There are M workers, and N jobs.
– The cost (or time) of having worker i to do job j is cij. How to
assign workers to complete the jobs so that the total cost is
minimized; or

– The return of having worker i to do job j is pij. How to assign

workers to complete the jobs so that the total return is
maximized.
 Assignment models are special cases of transportation models
where all flows are 0 or 1: each job requires 1 unit of inflow.

22
Network Formulation

Unit Shipping Cost: cij Demand nodes:

Supply nodes:
1
Workers Jobs
1
2

2
1 unit of flow from 3
Worker i to Job j if
and only if Worker All flows are
i is assigned to do either 0 or 1
job j
N
M

23
Example A: Assignment of Students to Companies
 Six HKUST Business School student teams have written business plans for
eight companies. Each plan was evaluated by internal professor-mentors and
given a score, which is given in follows. The school has to decide which
student team to do a presentation at which company.
 Each company will hear one and only one team's presentation. Each student
team will present at no more than two companies.

Scores of proposals

Company 1 Company 2 Company 3 Company 4 Company 5 Company 6 Company 7 Company 8

Team 1 35 39 63 78

Team 2 39 35 54 84 68
Team 3 69 73 61 81

Team 4 37 67 49 50 75
Team 5 45 53 78 53 89 87

Team 6 47 59 42 72 61 57 75
24
Network representation
2 Team 1 Company 1 1
35 39

39

2 Team 2 Company 2 1
35

Company 3 1
59
49
42

2 Team 6

75

Company 8 1

25
“Network” Formulation
 Imagine “supply” nodes for the student teams on the left: they
“supply” “Presentation”;
 “demand” nodes for the companies on the right: each
“demands” one “Presentation”
 An arc from team i to company j: if team i has a report for
company j;
 Unit “Reward” on an arc: score
 Flow: 1 or 0 (1 means the team is assigned to the company)

26
“Network” Formulation
 Constraints:
– Each company hears one proposal: inflow of each company node is
exactly 1.

– Each team can be assigned to at most two companies: outflow of any

team node to be at most 2.

 Objective: maximize the total scores

27
General Logistics Models
(Minimum Cost Network flow Models)

 Shipping products from some Suppliers, through

some Transshipment Points, to some
Demanders

2 5

1 7
4
8
3
6 28
General Logistics Models
 Nodes are generally categorized as
– Suppliers – a location that starts with a certain
supply

– Demanders – the opposite of a supplier, it

requires a certain amount to end up there

– Transshipment points – a location where

goods simply pass through
29
 Net inflow for any node: Total Inflow – Total Outflow
 Net outflow for any node: Total Outflow – Total Inflow

 Suppliers: Net Outflow > 0

 Demanders: Net Inflow > 0
 Transshipment points: Net Outflow = Net Inflow = 0

30
 Two types of constraints in general logistics models.
– The arc capacity constraints

– The flow balance constraints.

• For a supplier : net outflow <= supply capacity

• For a demander : net inflow >= demand (or net

inflow = demand )
• For a transshipment point: net inflow = net outflow
=0

31
Ex 5.4 Producing and Shipping Tomato
Products At RedBrand
 The RedBrand Company produces tomato products at three
plants.
 These products can be shipped directly to their two customers
or they can first be shipped to the company’s two warehouses
and then to the customers.
 At most 200 tons can be shipped between any two nodes.
 The cost of producing food at each plant is the same, so
RedBrand is concerned with minimizing the total shipping cost
incurred in meeting customer demands.

32
 Arc capacity = 200
Plant 1 Plant 2 Plant 3 Whse 1 Whse 2 Cust 1 Cust 2
Plant 1 $5.00 $3.00 $5.00 $5.00 $20.00 $20.00
Plant 2 $9.00 $9.00 $1.00 $1.00 $8.00 $15.00
Plant 3 $0.40 $8.00 $1.00 $0.50 $10.00 $12.00
Whse 1 $1.20 $2.00 $12.00
Whse 2 $0.80 $2.00 $12.00
Cust 1 $1.00
Cust 2 $7.00

33
Formulation
 We need to keep track of the following:

– Amount shipped along each arc of the network

– Net Outflow of the Supplier Node (Plant)

– Net Outflow or Net Inflow of each transhipment node

(warehouse)

– Net Inflow of each demander (Customer)

– Total shipping cost

34
Optimal
Solution

 Recall: The costs of shipping from plant 1 directly to warehouses

or customers are larger than shipping thru plant 3.
 Sensitivity analysis: How much effect does the arc capacity
have on optimal solution?

35
Ex 5.5 Shortest Walk
 Maude Jenkins, a 90-year old woman, is planning to walk across the state, west to
east, to gain support for a political cause she favors.
 She wants to travel the shortest distance to get from city 1 to city 10.

36
Formulation
 All we need to specify is that node 1 has a net supply of
1, node 10 has a net demand of 1, and all other nodes
are transshipment nodes.

37
Ex 5.7 Max Flow: Crude Oil Pipeline
 Excron Oil Company has a pipeline distribution network.
 Each node corresponds to a storage tank, and the numbers on
the arcs represent flow capacities (per hour).

 Excron wants to determine the maximum flow that can be sent

from tank 1 to tank 8 per hour.
38
 Optimal
Solution
…compare with
capacities

Most intermediate arcs

are not constraining.

 What if we double the capacity of all arcs leading into tank 8?

 Will this double the maximum flow?

39
Ex 5.7 Max Flow: Crude Oil Pipeline
 Suppose the company whishes to increase the maximum flow
from Tank 1 to Tank 8 by two units (per hour), through
increasing the capacities in some arcs. The cost of increasing
one unit of capacity in an arc is $100 millions. How could
Excron achieves the target with minimum cost?

40
Exercise:
9, 48
46,
47, 51, 53, 55, 62

Homework: Case: International Textile Company, LTD

•Due day: November 17

•Group: up to three

41