TheThe TransportationTransportation ModelModel
A special class of the LP problem
Deals with the situation in which commodities are shipped from sources to destinations
Objective is to determine the amounts shipped from each source to each destination
Need to minimise the total shipping cost
Need to satisfy both the supply limits and the demand requirements
TheThe TransportationTransportation ModelModel
Represent each of the m sources and each of the n destinations by a node
Link the sources to the destinations with directed edges (arcs)
This represents the routes between sources and destinations
Arc (i, j) joins source i to destination j
TheThe TransportationTransportation ModelModel
Arc (i, j) and carries two pieces of information
The transportation cost per unit, c _{i}_{j}
The amount shipped, x _{i}_{j}
The amount of supply at source i is a _{i}
The amount of demand at destination j is b _{j}
The objective is to determine the unknowns, x _{i}_{j}
Need to minimise total transportation cost
Need
all supply and demand
requirements
to
meet
TheThe TransportationTransportation ModelModel
Units of supply
sources
destinations
Units of demand
TheThe TransportationTransportation ModelModel
AUS Auto has three plants in Sydney, Perth and Melbourne
They also have two major distribution centers in Brisbane and Adelaide
The capacities of the three plants during the next quarter are 1000, 1500 and 1200 cars
The quarterly demand at the two distribution centers are 2300 and 1400 cars
TheThe TransportationTransportation ModelModel
The distances (in km) between the plants and distribution centers are (say)
Brisbane
Adelaide
Sydney 
1000 
2690 
Perth 
1250 
1350 
Melbourne 1275 850
TheThe TransportationTransportation ModelModel
The company that transports the cars charges 8 cents per km per car
The transportation cost per car on each route (to the nearest dollar) is then
Brisbane
(1)
Adelaide
(2)
Sydney (1) 
$80 
$215 
Perth (2) 
$100 
$108 
Melbourne (3) $102 $68
TheThe TransportationTransportation ModelModel
The LP model is
minimise z = 80x _{1}_{1} + 215x _{1}_{2} + 100x _{2}_{1} + 108x _{2}_{2} + 102x _{3}_{1} + 68x _{3}_{2} s. t.
x
x
x
x
x
_{1}_{1} + x _{1}_{2}
_{2}_{1} + x _{2}_{2}
_{3}_{1} + x _{3}_{2}
= 1000 (Sydney)
= 1500 (Perth)
= 1200 (Melbourne)
_{1}_{1} + x _{2}_{1} + x _{3}_{1} = 2300 (Brisbane)
_{1}_{2} + x _{2}_{2} + x _{3}_{2} = 1400 (Adelaide)
x 
_{i}_{j} ≥ 0, 
i 
= 1, 2, 3, 
j = 1, 2
TheThe TransportationTransportation ModelModel
The constraints are equations because
Total supply = 100 + 1500 + 1200 = 3700
Total demand is 2300 + 1400 = 3700
Can solve this using the simplex method !
The special structure of the constraints
allow us to solve the convenient way
problem in a more
TheThe TransportationTransportation ModelModel
The optimal solution is
1000
1500
1200
Sydney
1000
Perth
Melbourne
1200
Brisbane
Adelaide
2300
1400
Total transportation cost is $313,200
TheThe TransportationTransportation ModelModel
Solve the problem using a transportation tableau
Brisbane
Adelaide
Sydney 
80 
215 

^{x} 
11 
^{x} 
12 

Perth 
100 
108 

^{x} 
21 
^{x} 
22 

Melbourne 
102 
68 

^{x} 
31 
^{x} 
32 
Demand 2300
1400
Supply
1000
1500
1200
TheThe TransportationTransportation ModelModel
In the Aus Auto model, suppose that the Perth plant capacity is 1300 cars (instead of 1500)
This means that the total supply (3500) is less than the total demand (3700)
Part of the demand at Brisbane and Adelaide may not be filled
TheThe TransportationTransportation ModelModel
As demand exceeds supply, add a dummy source with a capacity of 200=37003500
This balances the transportation model
The c
is 0 because the plant does not exist
The c
may assume positive values as well
from the dummy plant to the destinations
ij
from the dummy source to the destinations
ij
e.g., to ensure that Adelaide will receive all its demand,
(penalty) to the entry from the
we assign a high c
dummy source to Adelaide
ij
TheThe TransportationTransportation ModelModel
The optimum solution of the balanced model is
Brisbane
Adelaide
80 
215 

Sydney 
1000 

100 
108 

Perth 
1300 

102 
68 

Melbourne 
1200 

0 
0 

Dummy 
200 

Demand 
2300 
1400 
_{S}_{u}_{p}_{p}_{l}_{y}
1000
1300
1200
200
The dummy plant ships 200 cars to Adelaide, which means that it will be 200 cars short !
TheThe TransportationTransportation ModelModel
Assume that the demand at Brisbane is only 1900 cars (not 2300)
Now supply exceeds the demand
In this case, we need to add a dummy distribution center to “receive” the surplus supply
Again, the c
’s to the dummy distribution center
ij
are 0, unless we wish to impose other conditions
e.g., we can require a factory to “ship out” completely
by assigning a high c the dummy destination
from the designated factory to
ij
TheThe TransportationTransportation ModelModel
The new model and its optimal solution is
Sydney
Perth
Melbourne
Demand
Supply
1000
1500
1200
The Perth plant will have a surplus of 400 cars
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The application of the transportation model is not limited to transporting commodities between geographical sources and destinations
A company C manufactures product P
Demand occurs during March to June
The company estimates the demand to be
100(March), 200(April), 180(May), 300(June)
The company’s production capacity varies
It estimates that it can produce
50 (March), 180(April), 280(May), 270(June)
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The production capacity and demand for the different months do not match
A current month’s demand may be satisfied in one of three ways
1. Current month’s production
cost per unit is $40
2. Surplus production in an earlier month incurs a holding cost of $0.50/unit/month
3. Surplus production in a later month incurs a penalty cost of $2/unit/month
Determine the optimal production schedule
NonNonTraditionalTraditional TransportationTransportation ModelsModels
Notice the following parallels between the elements of the productioninventory problem and the transportation model:
Transportation 1. Source i 
Productioninventory 1. Production period i 

2. 
Destination j 
2. Demand period j 
3. 
Supply amount at source i 
3. Production capacity of period i 
4. 
Demand at destination j 
4. Demand for period j 
5. 
Unit transportation cost from source i to destination j 
5. Unit cost (production+inventory +penalty) in period i for period j 
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The unit “transportation” cost from period i to period j is computed as
c _{i}_{j} =
Production cost in i, i =j
Production cost in i + holding cost from i to j, i < j Production cost in i + penalty cost from i to j, i > j
e.g.
c _{1}_{1} = $40.00
c _{2}_{4} = $40.00 + ($0.50 + $0.50) = $41.00
c _{4}_{2} = $40.00 + ($2.00 + $2.00) = $44.00
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The resulting transportation model is
1 
2 
3 
4 
Capacity 

1 
$40 
$40.5 
$41 
$41.5 
50 
2 
$42 
$40 
$40.5 
$41 
180 
3 
$44 
$42 
$40 
$40.5 
280 
4 
$46 
$44 
$42 
$40 
270 
Demand 
100 
200 
180 
300 
NonNonTraditionalTraditional TransportationTransportation ModelsModels
supply
Supply
Demand
100
200
180
300
demand
NonNonTraditionalTraditional TransportationTransportation ModelsModels
A saw mill prepares different types of wood
that range from soft pine to hard oak according
to a weekly schedule
Depending on the type of wood being milled, the demand for sharp blades varies from day to day according to the following 1week (7day) data:
Day 
Mon Tues Wed Thur Fri 
Sat 
Sun 

Demand 
24 
12 
14 
20 
18 
14 
22 
(blades)
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The mill can satisfy the daily demand in the following manner:
1.
Buy new blades at the cost of $12 a blade
2. Use an overnight sharpening service at the cost of $6 a blade, or a slow 2day service at the cost of $3 a blade.
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The
transportation model
The 7 destinations represent the 7 days of the week
The 8 sources are defined as follows:
a
situation
be
represented
can
as
Source 1: buying new blades,
Sources 2 to 8: the 7 days of the week
NonNonTraditionalTraditional TransportationTransportation ModelsModels
The amount of supply for each of these sources equals the number of used blades at the end of the associated day
For example, source 2 (i.e., Monday) will have a supply of used blades equal to the demand for Monday
The unit transportation cost for the model could be $12, $6, or $3, depending on whether the blade is supplied from new blades, overnight sharpening, or 2day sharpening
NonNonTraditionalTraditional TransportationTransportation ModelsModels
Notice that the overnight service means that used blades sent at the _{e}_{n}_{d} of day i will be in use at the start of day i + 1 or day i + 2.
The 2day service means that used blades sent at the end of day i will be in use at the start of day i + 3 or any day thereafter
The “disposal” column is a dummy destination that receives the blades left unsharpened at the end of a day
NonNonTraditionalTraditional TransportationTransportation ModelsModels
1 New
2 Mo
3 Tu
4 We
5 Th
6 Fr
7 Sa
8 Su
1 
2 
3 
4 
5 
6 
7 
8 
Mo 
Tu 
We 
Th 
Fr 
Sa 
Su 
Dis 
$12 
$12 
$12 
$12 
$12 
$12 
$12 
$0 
24 
2 
98 

M 
$6 
$6 
$3 
$3 
$3 
$3 
$0 
10 
8 
6 

M 
M 
$6 
$6 
$3 
$3 
$3 
$0 
6 
6 

M 
M 
M 
$6 
$6 
$3 
$3 
$0 
14 

M 
MMM 
$6 
$6 
$3 
$0 

12 
8 

M 
M 
M 
M 
M 
$6 
$6 
$0 
14 
4 

M 
MM 
M 
MM 
$6 
$0 

14 

M 
M 
M 
M 
M 
M 
M 
$0 
22 

24 
12 
14 
20 
18 
14 
22 
124 
124
24
12
14
20
18
14
22
NonNonTraditionalTraditional TransportationTransportation ModelsModels
SharpeningService
Period 
New 
Overnight 
2Day 
Disp 
Mo 
24(Mo) 
10(Tu)+8(W) 
6(Th) 
0 
Tu 
2(Tu) 
6(We) 
6(Fr) 
0 
We 
0 
14(Th) 
0 
0 
Th 
0 
12(Fr) 
8(Su) 
0 
Fr 
0 
14(Sa) 
0 
4 
Sa 
0 
14(Su) 
0 
0 
Su 
0 
0 
0 
22 
NonNonTraditionalTraditional TransportationTransportation ModelsModels
At the start of Monday, the company buys 24 new blades
At the end of Mon, they will have 24 used blades
18 sent to overnight service, 6 to 2day service
Of the 18 overnight blades,
10 will be used on Tuesday, and 8 on Wednesday
The six 2day blades will be used on Thursday etc.
The disposal column shows the # of used blades that will be left unsharpened at the end of a given day
TheThe TransportationTransportation AlgorithmAlgorithm
The algorithm follows the exact steps of the simplex method
instead of using the regular simplex tableau, we take advantage of the special structure of the transportation model to present the algorithm in a more convenient form
We use the following numeric example to present the details of the algorithm
TheThe TransportationTransportation AlgorithmAlgorithm
A transport company ships truckloads of grain from three silos to four mills
The supply and demand is in #truckloads The c _{i}_{j} ’s (topright corner of each box) are in $’00s
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
TheThe TransportationTransportation AlgorithmAlgorithm
The purpose of the model is to determine the minimum cost shipping schedule between the silos and the mills
This is equivalent to determining the quantity x _{i}_{j} shipped from silo i to mill j
(i = 1, 2, 3; j = 1, 2, 3, 4)
TheThe TransportationTransportation AlgorithmAlgorithm
The steps of the transportation algorithm are exact parallels of the simplex algorithm
1. Determine a starting basic feasible solution
2. If the optimality condition is satisfied, stop. Otherwise, use the optimality condition of the simplex method to determine the entering variable from amongst all the nonbasic variables Go to step 3
3. Use the feasibility condition of the simplex method to determine the leaving variable from among all the current basic variables and find the new basic solution Return to step 2
TheThe TransportationTransportation AlgorithmAlgorithm
The special structure of the transportation problem allows securing a nonartificial starting basic solution using one of three methods:
1. Northwestcorner method 2. Leastcost method 3. Vogel approximation method
The difference the methods is the “quality” of the starting basic solution they produce
In general, 3. yields the best and 1. yields the worst
However 1. involves the least computations.
TheThe NorthNorthWestWestCornerCorner MethodMethod
Step 1.
Step 2.
Step 3.
, the northwest corner of the (remaining) tableau Adjust the associated amounts of supply and demand by subtracting the allocated amount If all units have been allocated stop; else ifthe supply in row i is satisfied AND the demand in column j is satisfied, remove EITHER row i OR row j (but not BOTH)
Allocate as much as possible to the variable, x _{i}_{j}
in
else if 
the supply in row i is satisfied 
else if 
remove row i the demand in column j is satisfied remove column j 
Go to Step 1.
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Update variable x _{1}_{1} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove column 1
Supply
10
25
10
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
10
25
10
Update variable x _{1}_{2} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
10
25
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
10
25
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove row 1
Supply
0
25
10
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Update variable x _{2}_{2} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove column 2
Supply
0
20
10
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
20
10
Update variable x _{2}_{3} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
20
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
20
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove column 3
Supply
0
5
10
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
5
10
Update variable x _{2}_{4} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
5
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
5
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove row 2
Supply
0
0
10
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
0
10
Update variable x _{3}_{4} (NW corner)
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
0
10
Allocate as much as possible
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
0
10
Update supply and demand
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
0
0
All units have been allocated; stop.
TheThe NorthNorthWestWestCornerCorner MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
The starting basic solution is given as
x _{1}_{1} = 5, x _{1}_{2} = 10, x _{2}_{2} = 5, x _{2}_{3} = 15, x _{2}_{4} = 5, x _{3}_{4} = 10
The associated cost of the schedule is
z = 5x10 + 10x2 + 5x7 + 15x9 + 5x20 +10x18 = $520
TheThe LeastLeastCostCost MethodMethod
Concentrate on the “cheapest” cells
Start by assigning as much as possible to the cell with the smallest unit cost (ties are broken arbitrarily)
We then adjust the amounts of supply and demand and remove the satisfied row or column accordingly
If both a row and a column are satisfied simultaneously, only one is removed
Repeat until all units have been allocated
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Update variable x _{1}_{2} (cheapest = 2)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Allocate as much as possible
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
15
25
10
Update supply and demand
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Arbitrarily remove column 2
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Update variable x _{3}_{1} (cheapest = 4)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Allocate as much as possible
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
10
Update supply and demand
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove column 1
Supply
0
25
5
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
5
Update variable x _{2}_{3} (cheapest = 9)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
5
Allocate as much as possible
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
25
5
Update supply and demand
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove column 3
Supply
0
10
5
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Update variable x _{1}_{4} (cheapest = 11)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Allocate as much as possible
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Update supply and demand
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove row 1
Supply
0
10
5
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Update variable x _{3}_{4} (cheapest = 18)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Allocate as much as possible
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
5
Update supply and demand
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Remove row 3
Supply
0
10
0
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Silo 3
Demand
Supply
0
10
0
Update variable x _{2}_{4} (cheapest = 20)
TheThe LeastLeastCostCost MethodMethod
Silo 1
Silo 2
Much more than documents.
Discover everything Scribd has to offer, including books and audiobooks from major publishers.
Cancel anytime.