M C S D 11 3 3 O p e r a t i o n s R e s e a r c h & O p t i m i z a t i o n

CHAPTER 7: ILP -
TRANSPORTATION MODEL
(PART 1)

[email protected](S1)

UTM JOHOR BAHRU

 The Transportation Model
§ A product is transported from a few sources to several destinations at the minimum
possible cost.
§ Each source can supply a fixed number of units of the product, and each
destination has a fixed demand for the product.
§ The linear programming model has constraints for supply at each source and
demand at each destination.
§ All constraints are equalities in a balanced transportation model where supply
equals demand.
§ Constraints contain inequalities in unbalanced models where supply does not equal
demand.

2
 The Transportation Model
§ There are m sources and n destinations, each represented by a node. The arcs (i, j)
represent the routes linking the sources and the destinations.
§ To develop a model of a transportation problem, it is necessary to have the following
information:
o Supply quantity (capacity) at source i, si, ,.
o Demand quantity of each destination j, dj.
o Arc (i, j) joining source i to destination j carries two pieces of information: the
transportation cost per unit, cij, and the amount shipped, xij.
§ The objective of the model is to minimize the total transportation cost while satisfying
all the supply and demand restrictions.

3
 The Transportation Model
Schematic of a Transportation Problem

4
 The Transportation Model
Balanced Transportation Problem

Model Formulation =>

𝑥𝑖𝑗 for all i and j is said to be balanced transportation problem when total supply from all the sources is
equal to the total demand in all destinations. Otherwise, problem is said to be unbalanced transportation
problem.

5
 The Transportation Model

§ Transportation problem is considered such an important special type of linear

programming problem.
§ For many applications, the supply and demand quantities in the model (the si and dj)
have integer values, and implementation will require that the distribution quantities
(the xij) also have integer values.
§ The solution procedure deals only with basic feasible solutions, so it automatically will
obtain an integer optimal solution for this case. Therefore, it is unnecessary to add a
constraint to the model that the xij must have integer values.

6
 The Transportation Model
Unbalanced Transportation Problem

The unit transportation cost for the dummy supply variable and dummy
demand variable are assigned zero values, because no shipment is made.

7
 Example 1

How many tons of wheat to transport from each grain elevator to each mill on a
monthly basis in order to minimize the total cost of transportation?

Grain Elevator Supply Mill Demand

1. Kansas City 150 A. Chicago 200
2. Omaha 175 B. St. Louis 100
3. Des Moines 275 C. Cincinnati 300
Total 600 tons Total 600 tons

Transport Cost from Grain Elevator to Mill ($/ton)

Grain Elevator A. Chicago B. St. Louis C. Cincinnati
1. Kansas City $6 $8 $ 10
2. Omaha 7 11 11
3. Des Moines 4 5 12

8
 Example 1 (cont’d)

200

Network of Transportation Routes for Wheat Shipments

9
 Example 1 (cont’d)
Model Formulation
Transport Cost from Grain Elevator to Mill ($/ton)
Grain Elevator A. Chicago B. St. Louis C. Cincinnati
1. Kansas City $6 $8 $ 10
2. Omaha 7 11 11
3. Des Moines 4 5 12

Minimize Z = 6x1A + 8x1B + 10x1C + 7x2A + 11x2B + 11x2C + 4x3A + 5x3B + 12x3C
subject to:
x1A + x1B + x1C = 150 (Supply from Kansas City)
x2A + x2B + x2C = 175 (Supply from Kansas Omaha)
x3A + x3B + x3C = 275 (Supply from Kansas Des Moines)
x1A + x2A + x3A = 200 (Demand at St. Louis)
x1B + x2B + x3B = 100 (Demand at Cincinnati)
x1C + x2C + x3C = 300
xmn ³ 0
xmn = tons of wheat from each grain elevator, m, m = 1, 2, 3, to each mill n, n = A,B,C

10
 Example 1 (cont’d)

Transportation Model Solution Using Excel Solver

Transport Cost from Grain Elevator to Mill ($/ton)

Grain Elevator A. Chicago B. St. Louis C. Cincinnati
1. Kansas City $6 $8 $ 10
2. Omaha 7 11 11
3. Des Moines 4 5 12

11
 Example 1 (cont’d)
Transportation Model Solution Using Excel Solver

12
 Example 1 (cont’d)

Transportation Model Solution Using Excel Solver

13
 Example 1 (cont’d)

Transportation Model Solution (summary)

14
 Example 1 (cont’d)
Transportation Model Solution Using Python
1. Import library (PuLP) and define node and capacities
#Import PuLP modeler functions
from pulp import *

#Creates a list of all the supply nodes

GrainE = ["1", "2","3"]

#Creates a dictionary for the number of units of supply for each supply node
supply = {"1": 150,
"2": 175,
"3": 275}
#Creates a list of all demand nodes
Mills = ["A", "B", "C"]
#Creates a dictionary for the number of units of demand for each demand node
demand = {"A":200,
"B":100,
"C":300 }

15
 Example 1 (cont’d)
Transportation Model Solution Using Python (cont’d)
2. Define cost shipping
The cost data is then inputted into a list, containing the costs of shipping from the supply to demand
nodes. Note here that the commenting and structure of the code makes the data appear as a table for
easy editing. The Grain Elevator and Mills lists (Supply and Demand nodes) are added to make a large list
(of all nodes) and inputted into PuLPs makeDict function.

costs = [#Grain Elevator

#A B C
[6,8,10],# 1 Mills
[7,11,11], #2
[4,5,12] #3
]

# The cost data is made into a dictionary

costs = makeDict([GrainE,Mills],costs,0)

16
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

3. Create the variable to contain the problem data.
Install using LpProblem function.

# Creates the 'prob' variable to contain the problem data

prob = LpProblem("Grain Distribution Problem",LpMinimize)

It has two parameters,

I. The arbitrary name of this problem (as a string),
II. Second parameter being either LPMinimize or LpMaximize depending on the type of LP to solve.

17
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

4. Create Routes.
A list of tuples is created containing all the possible route for the transportation.

# Creates a list of tuples containing all the possible routes for transport
Routes = [(i,j) for i in GrainE for j in Mills]

5. Create vars dictionary.

A dictionary called vars is created which contains the LP variables. The reference keys to the dictionary
are the Grain Elevator name, then the mills name([“1”][“A”]) , and the data is Route_Tuple.
(e.g. [“1”][“A”]: Route_1_A). The lower limit of zero is set, the upper limit of None is set, and the variables
are defined to be Integers.

# A dictionary called ‘vars' is created to contain the referenced variables(the routes)

vars = LpVariable.dicts("Route",(GrainE,Mills),0,None,LpInteger)

18
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

6. Set the objective function.
The objective function is added to the variable prob using a list comprehension.
Since vars and costs are now dictionaries (with further internal dictionaries), they can be used
as if they were tables, as for (i,j) in Routes will cycle through all the combinations/arcs.

# The objective function is added to 'prob' first

prob += lpSum([vars[i][j]*costs[i][j] for (i,j) in Routes]), "Sum_of_Transporting_Costs"

19
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

7. Set the supply constraints.
§ The supply constraints are added using a normal for loop and a list comprehension.
§ Supply Constraints: For each Grain Elevator in turn, the values of the decision variables (number of
grain cases on arc) to each of the mills is summed, and then constrained to being less than or
equal to the supply max for that grain elevator.
§ “Add constraints ‘+=’ operator again to add more data to the variable prob.

# The supply maximum constraints are added to prob for each supply node (Grain Elavator)
for i in GrainE:
prob += lpSum([vars[i][j] for j in Mills]) ==supply[i], "Sum_of_Products_out_of_GrainE_%s"%i

20
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

8. Set the demand constraints.
§ The demand constraints are added using a normal for loop and a list comprehension.
§ Demand Constraints: For each mill in turn, the values of the decision variables (number on arc) from
each of the grain Elavator is summed, and then constrained to being greater than or equal to the
demand minimum.
§ “Add constraints ‘+=’ operator again to add more data to the variable prob.

# The demand minimum constraints are added to prob for each demand node (Mills)
for j in Mills:
prob += lpSum([vars[i][j] for i in GrainE]) ==demand[j], "Sum_of_Products_into_Mill%s"%j

21
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

9. Add WriteLP function
§ WriteLP function can be used to copy the information into a .lp file into the directory that the code-
block is running from.
§ There is no assignment operator (such as an equal's sign) on this line. This is because the
function/method WriteLP called is being performed to the variable/object prob.
§ The dot . between the variable/object and the function/method is important and is seen frequently in
Object Oriented software

# The problem data is written to an .lp file

prob.writeLP("GrainDistributionProblem.lp")

22
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

10. Add solve() function
§ The LP is solved using the solver that PuLP chooses. The input brackets after solve( )are left empty in
this case, however they can be used to specify which solver to use.

# The problem is solved using PuLP's choice of Solver

prob.solve()

23
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

11 . Display output
§ Now the results of the solver call can be displayed as output.
§ Request the status of the solution, which can be one of “Not Solved”, “Infeasible”,
“Unbounded”, “Undefined” or “Optimal”.
§ The value of prob is returned as an integer, which must be converted to its significant text
meaning using the lpStatus dictionary.

# The status of the solution is printed to the screen

print ("Status:", LpStatus[prob.status])

24
 Example 1 (cont’d)

Transportation Model Solution Using Python (cont’d)

11 . Display output (cont’d)
§ The variables and their resolved optimum values can now be printed to the screen.
§ The for loop is used to called all the problem variable names. Prints each variable name, followed by an
equal's sign, followed by its optimum value. The optimised objective function value is printed to the screen,
using the value (prob.objective) function

# Each of the variables is printed with it's resolved optimum value

for v in prob.variables():
print(v.name, "=", v.varValue)
# The optimised objective function value is printed to the screen
print ("Total Cost of Transportation = ", value(prob.objective))

25
 Exercise

A Cola Company has two warehouses from which it distributes cola to five carefully
chosen stores. At the start of every week, each stores sends an order to the Cola
Company head office, the Cola is then dispatched from the appropriate
warehouse to the store. The Cola company would like to have an interactive
computer program which they can run week by week to tell them which
warehouse should supply which store so as to minimize the costs of the whole
operation.
Suppose that at the start of a given week the company has 1000 cases of Cola
cans at warehouse A, and 4000 cases of Cola cans at warehouse B, and that the
stores(1,2,3,4,5) require 500, 900, 1800, 200, and 700 cases, respectively. Which
warehouse should supply which store?

26
 Exercise

Cost for warehouse A and B to Store (1,2,3,4,5)

27