Transportation problem

Example
P&T Company produces canned peas.
Peas are prepared at three canneries (Bellingham,
Eugene and Albert Lea).
Shipped by truck to four distributing warehouses
(Sacramento, Salt Lake City, Rapid City and
Albuquerque). 300 truckloads to be shipped.
 Problem: minimize the total shipping cost.

TRANSPORTATION AND
ASSIGNMENT PROBLEMS

Joo Miguel da Costa Sousa / Alexandra Moutinho

P&T Company problem

127

Shipping data for P&T problem

Shipping cost per truckload in
Warehouse

Cannery

464

513

654

867

75

352

416

690

791

125

995

682

388

685

100

80

65

70

85

300

Allocation

Joo Miguel da Costa Sousa / Alexandra Moutinho

128

Network representation

Output

Joo Miguel da Costa Sousa / Alexandra Moutinho

129

Formulation of the problem

Minimize Z = 464 x + 513 x + 654 x + 867 x + 352x + 416 x
11

12

13

14

21

+ 690 x + 791 x + 995 x + 682 x + 388 x + 685 x

23

subject to

24

31

32

33

22

34

x11 + x 12 + x 13 + x 14 = 75
x21 + x 22 + x23 + x24 = 125
x 31 + x 32 + x 33 + x 34 = 100
x11 + x21 + x 31 = 80
x 12 + x22 + x 32 = 65
x 13 + x23 + x 33 = 70
x14 + x24 + x 34 = 85

and x ij 0
Joo Miguel da Costa Sousa / Alexandra Moutinho

130

(i = 1,2,3,4; j = 1,2,3,4)

Joo Miguel da Costa Sousa / Alexandra Moutinho

131

Constraints coefficients for P&T Co.

Transportation problem model

Coefficient of:
x 11

x12

x 13

x 14

A = 1

x21

x22

x23

x24

x 31

1
1
1

x 33

1
1

x32

1
1

1
1

x 34

Joo Miguel da Costa Sousa / Alexandra Moutinho

Cannery

constraints

Warehouse

constraints

132

Transportation problem model

134

Transportation problem model

General problem

Truckload of canned peas

Units of a commodity

Three canneries

m sources

Four warehouses

n destinations

Output from cannery i

Supply si from source i

Allocation to warehouse j

Demand dj at destination j

Shipping cost per truckload from

cannery i to warehouse j

Cost cij per unit distributed from source i

to destination j

Joo Miguel da Costa Sousa / Alexandra Moutinho

133

The feasible solutions property: a transportation

problem has feasible solution if and only if
m

i =1

j =1

si = d j
If the supplies represent maximum amounts to be
distributed, a dummy destination can be added.
Similarly, if the demands represent maximum
amounts to be received, a dummy source can be
added.
Joo Miguel da Costa Sousa / Alexandra Moutinho

135

Parameter table for transportation prob.

The cost assumption: the cost of distributing units

from any source to any destination is directly
proportional to the number of units distributed.
This cost is the unit cost of distribution times the
number of units distributed.
Integer solution property: for transportation
problems where every si and dj have an integer value,
all basic variables in every BF solution also have
integer values.
Joo Miguel da Costa Sousa / Alexandra Moutinho

Prototype example

Transportation problem model

Each source has a certain supply of units to distribute

to the destinations.
Each destination has a certain demand for units to be
received from the source .
Requirements assumption: Each source has a fixed
supply of units, which must be entirely distributed to
the destinations. Similarly, each destination has a
fixed demand for units, which must be entirely
received from the sources.
Joo Miguel da Costa Sousa / Alexandra Moutinho

Transportation problem: distributes any commodity

from any group of sources to any group of
destinations, minimizing the total distribution cost.

136

Cost per unit distributed

Source

Demand

1
2

1
c11
c21

cm1
d1

Joo Miguel da Costa Sousa / Alexandra Moutinho

Destination
2

c12

c22

cm2

d2

n
c1n
c2n

cmn
dn

Supply
s1
s2

sm

137

Transportation problem model

Network representation

The model: any problem fits the model for a

transportation problem if it can be described by a
parameter table and if the satisfies the requirements
assumptions and the cost assumption.
The objective is to minimize the total cost of
distributing the units.
 Some problems that have nothing to do with
transportation can be formulated as a transportation
problem.
Joo Miguel da Costa Sousa / Alexandra Moutinho

138

Joo Miguel da Costa Sousa / Alexandra Moutinho

Coefficient constraints

Mathematical formulation of the problem

 Z: total distribution cost
 xij : number of units distributed from source i to destination j
m

i =1 j =1

subject to
n

ij

= si , for i = 1,2, , m,

j =1
m

ij

Coefficient of:
x 11
1

A =
1

Minimize Z = c ij x ij ,

= d j , for j = 1,2, , n,

j =1

139

x 12 ... x 1 n
1

x 21

x 22 ... x2n

... x m1

x m2 ... x mn

1
1

1
1

1
1

Supply

constraints

Demand

constraints

and x ij 0, for all i and j.

Joo Miguel da Costa Sousa / Alexandra Moutinho

140

Example: solving with Excel

Joo Miguel da Costa Sousa / Alexandra Moutinho

Joo Miguel da Costa Sousa / Alexandra Moutinho

141

Example: solving with Excel

142

Joo Miguel da Costa Sousa / Alexandra Moutinho

143

The transportation simplex method

Transportation simplex tableau (TST)

Destination

Version of the simplex called transportation simplex

method.
Problems solved by hand can use a transportation
simplex tableau.
Dimensions:

Source

144

c21

c12
c22

cm1

Demand

c1n

c2n

cm2

d1

S2

cmn

ui

s1

d2

Supply

sn

dn

Z=

vj
If xij is a nonbasic
variable:

If xij is a basic
variable:

Additional
information to be
added to each cell:

cij

cij
cij-ui-vj

xij

Joo Miguel da Costa Sousa / Alexandra Moutinho

The transportation simplex method

145

The transportation simplex method

Initialization: construct an initial BF solution. To

begin, all source rows and destination columns of the
TST are initially under consideration for providing a
basic variable (allocation).
1. From the rows and columns still under consideration,
select the next basic variable (allocation) according to one
of the criterions:
 Northwest corner rule;
 Vogels approximation method;
 Russells approximation method.

Joo Miguel da Costa Sousa / Alexandra Moutinho

c11

 For a transportation problem with m sources and n

destinations, simplex tableau has m + n + 1 rows and (m +
1)(n + 1) columns.
 The transportation simplex tableau has only m rows and n
columns!
Joo Miguel da Costa Sousa / Alexandra Moutinho

146

2. Make that allocation large enough to exactly use up the

remaining supply in its row or the remaining demand in its
column (whatever is smaller).
3. Eliminate that row or column (whiever had the smaller
remaining supply or demand) from further consideration. (If
the row or column have the same remaining supply and
demand, arbitrarily select the row as the one to be
eliminated. The column will be used later to provide a
degenerate basic variable, i.e., a vircled allocation of zero.)

Joo Miguel da Costa Sousa / Alexandra Moutinho

The transportation simplex method

147

Metro Water District example

4. If only one row or one column remains under consideration,

then the procedure is completed by selecting every
remaining variable (i.e., those variables that were neither
previously selected to be basic nor eliminated from
consideration by eliminating their row or column) associated
with that row or column to be basic with the only feasible
allocation. Otherwise return to step 1.

Initial BF solution from the Nortwest Corner Rule:

Destination
1

Source

 Go to the optimality test.

16

20

13

14

14

13

19

19

20

16

2
3
4(D)

Demand

2
30

30

20

22
60

19

10

23

30

ui

50

15
30

70

Supply

17

60

10

50

60

50
50
Z=2,470+10M

vj

Joo Miguel da Costa Sousa / Alexandra Moutinho

148

Joo Miguel da Costa Sousa / Alexandra Moutinho

149

The transportation simplex method

The transportation simplex method

Optimality test: derive ui and vj by selecting the row

having the largest number of allocations, setting its
ui=0, and then solving the set of equations cij = ui+ vj
for each (i,j) such that xij is basic. If cij ui vj 0 for
every (i,j) such that xij is nonbasic, then the current
solution is optimal, so stop. Otherwise, go to an
iteration.

Iteration:
1. Determine the entering basic variable: select the
nonbasic variable xij having the largest (in absolute
terms) negative value of cij ui vj.

Joo Miguel da Costa Sousa / Alexandra Moutinho

Joo Miguel da Costa Sousa / Alexandra Moutinho

150

The transportation simplex method

2. Determine the leaving basic variable: identify the

chain reaction required to retain feasibility when the
entering basic variable is increased. From the donor
cells, select the basic variable having the smallest
value.
151

Assignment problem

Iteration:
3. Determine the new BF solution: add the value of the
leaving basic variable to the allocation for each
recipient cell. Subtract this value from the allocation
for each donor cell.
4. Apply the optimality test.

Special type of linear programming where assignees

are being assigned to perform tasks.

Example: employees to be given work assignments.
Assignees can be machines, vehicles, plants or even
time slots to be assigned tasks.

Joo Miguel da Costa Sousa / Alexandra Moutinho

Joo Miguel da Costa Sousa / Alexandra Moutinho

152

Assumptions of assignment problems

1. The number of assignees and the number of tasks
are the same (if necessary, use dummy assignees
and dummy tasks so that nassignees= ntasks= n).
2. Each assignee is to be assigned to exactly one task.
3. Each task is to be performed by exactly one
assignee.
4. There is a cost cij associated with assignee i
performing task j (i,j = 1, 2, , n).
5. The objective is to determine how all n assignments
should be made to minimize the total cost.
Joo Miguel da Costa Sousa / Alexandra Moutinho

154

153

Prototype example
 Job Shop Company has purchased three new machines of
different types, and there are four different locations in the
shop where a machine can be installed.
 Objective: assign machines to locations.
Cost per hour of material handling (in )
Location
1
2
3
1
13
16
12
Machine
2
15

13
3
5
7
10
Joo Miguel da Costa Sousa / Alexandra Moutinho

4
11
20
6
155

Assignment problem model

Formulation as an assignment problem

 Using a dummy machine and an extremely large cost c22=M:
Task (Location)
1

if assignee i performs task j ,

if not.
n

Minimize Z = c ij x ij ,
subject to

13

16

12

11

15

13

20

10

4(D)

Assignee
(Machine)

1
 Decision variables x ij =
0

ij

= 1, for i = 1,2, , n,

ij

= 1, for j = 1,2, , n,

j =1
n

x
j =1

 Optimal solution: machine 1 to location 4, machine 2 to

location 3 and machine 3 to location 1 (total cost of 29 per
hour)
Joo Miguel da Costa Sousa / Alexandra Moutinho

i =1 j =1
n

and x ij 0, for all i and j

( x ij binary , for all i and j )
156

Assignment vs. Transportation prob.

Joo Miguel da Costa Sousa / Alexandra Moutinho

Still a linear
programming
problem?
157

Network representation

Assignment problem is a special type of

transportation problem where sources assignees
and destinations tasks and:
 #sources m = # destinations n;
 Every supply si = 1;
 Every demand dj = 1.

Due to the integer solution property, since si and dj are

integers, every BF solution is an integer solution for an
assignment problem. We may delete the binary
restriction and obtain a linear programming problem!
Joo Miguel da Costa Sousa / Alexandra Moutinho

158

Cost per unit distributed

1
2
Source

m=n
Demand

Joo Miguel da Costa Sousa / Alexandra Moutinho

Destination
2

c12

c22

cn2

n
c1n
c2n

cnn
1

159

Concluding remarks

Parameter table as transportation prob.

1
c11
c21

cn1
1

Joo Miguel da Costa Sousa / Alexandra Moutinho

Supply
1
1

160

Besides the general simplex and the transportation

simplex methods, a special algorithm for the
assignment problem is the Hungarian algorithm
(more efficient).
Streamlined algorithms were developed to explore the
special structure of some linear programming
problems: transportation and assignment problems.
Transportation and assignment problems are special
cases of minimum cost flow problems.
Network simplex method solves this type of problems.
Joo Miguel da Costa Sousa / Alexandra Moutinho

161