0% found this document useful (0 votes)
12 views164 pages

Lecture 06 - SpeciallyStructuredLPP - Transportation&Transshipment

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views164 pages

Lecture 06 - SpeciallyStructuredLPP - Transportation&Transshipment

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 164

Specially Structured Linear Programming Problems

Transportation, Transshipment and Assignment Problems

December 1, 2023
Introduction
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
▶ Availability at Source node i is ai
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
▶ Availability at Source node i is ai
▶ Demand at Destination node j is bj
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
▶ Availability at Source node i is ai
▶ Demand at Destination node j is bj
▶ Availability Constraints are given by
Xn
xij ≤ ai , ∀ i = 1, 2, . . . m (1)
j=1
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
▶ Availability at Source node i is ai
▶ Demand at Destination node j is bj
▶ Availability Constraints are given by
Xn
xij ≤ ai , ∀ i = 1, 2, . . . m (1)
j=1

▶ Demand Constraints are given by


Xm
xij ≥ bj , ∀ j = 1, 2, . . . n (2)
i=1

▶ All xij ≥ 0
Formulation of Transportation Models
▶ xij be the quantity to be transported from Node i to j
▶ Availability at Source node i is ai
▶ Demand at Destination node j is bj
▶ Availability Constraints are given by
Xn
xij ≤ ai , ∀ i = 1, 2, . . . m (1)
j=1

▶ Demand Constraints are given by


Xm
xij ≥ bj , ∀ j = 1, 2, . . . n (2)
i=1

▶ All xij ≥ 0
▶ Objective Function is m X
n
X
min Z = cij xij (3)
i=1 j=1
Formulation of Transportation Models
Pm Pn
▶ When i=1 ai ≥ j=1 bj then only it will be a successful TP
Formulation of Transportation Models
▶ When m
P Pn
i=1 ai ≥ j=1 bj then only it will be a successful TP
▶ When the above condition is not met, then not all Demands get satisfied
Formulation of Transportation Models
▶ When m
P Pn
i=1 ai ≥ j=1 bj then only it will be a successful TP
▶ When the above condition is not met, then not all Demands get satisfied
▶ Assuming cij ≥ 0 is a valid assumption for this
Formulation of Transportation Models
When i=1 ai ≥ nj=1 bj then only it will be a successful TP
Pm

P

▶ When the above condition is not met, then not all Demands get satisfied
▶ Assuming cij ≥ 0 is a valid assumption for this
▶ If cij ≥ 0, the sources will send exactly the demanded quantity because, sending
more than required will cost more
Formulation of Transportation Models
When i=1 ai ≥ nj=1 bj then only it will be a successful TP
Pm

P

▶ When the above condition is not met, then not all Demands get satisfied
▶ Assuming cij ≥ 0 is a valid assumption for this
▶ If cij ≥ 0, the sources will send exactly the demanded quantity because, sending
more than required will cost more
▶ So, all the demand constraints end up becoming Equality i.e.
Xm
xij = bj , ∀ j = 1, 2, . . . n (4)
i=1
Formulation of Transportation Models
When i=1 ai ≥ nj=1 bj then only it will be a successful TP
Pm

P

▶ When the above condition is not met, then not all Demands get satisfied
▶ Assuming cij ≥ 0 is a valid assumption for this
▶ If cij ≥ 0, the sources will send exactly the demanded quantity because, sending
more than required will cost more
▶ So, all the demand constraints end up becoming Equality i.e.
Xm
xij = bj , ∀ j = 1, 2, . . . n (4)
i=1
Pm Pn
▶ If somehow, the Total Supply becomes equal to Total Demand i.e. i=1 ai = j=1 bj
then the Availability Constraints also become equality i.e.
Xn
xij = ai , ∀ i = 1, 2, . . . m (5)
j=1
Formulation of Transportation Models
When i=1 ai ≥ nj=1 bj then only it will be a successful TP
Pm

P

▶ When the above condition is not met, then not all Demands get satisfied
▶ Assuming cij ≥ 0 is a valid assumption for this
▶ If cij ≥ 0, the sources will send exactly the demanded quantity because, sending
more than required will cost more
▶ So, all the demand constraints end up becoming Equality i.e.
Xm
xij = bj , ∀ j = 1, 2, . . . n (4)
i=1
Pm Pn
▶ If somehow, the Total Supply becomes equal to Total Demand i.e. i=1 ai = j=1 bj
then the Availability Constraints also become equality i.e.
Xn
xij = ai , ∀ i = 1, 2, . . . m (5)
j=1

▶ If m
P Pn
i=1 ai = j=1 bj then its a Balanced TP, otherwise it is Unbalanced TP which
has inequalities
Transport Tableau

D1 D2 ... Dn Supply, ai
c11 c12 c1n
S1 x11 x12 ... x1n a1

c21 c22 c2n


S2 x21 x22 ... x2n a2

... ... ... .. ... ...


.
cm1 cm2 cmn
Sm xm1 xm2 ... xmn am
n
X m
X
Demands, bj b1 b2 ... bn bj = ai
j=1 i=1
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
▶ Vogel’s Approximation Method
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
▶ Vogel’s Approximation Method
▶ Finding Optimal Solution
▶ Stepping Stone Method
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
▶ Vogel’s Approximation Method
▶ Finding Optimal Solution
▶ Stepping Stone Method
▶ Modified Distribution (MODI) Method or uv method
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
▶ Vogel’s Approximation Method
▶ Finding Optimal Solution
▶ Stepping Stone Method
▶ Modified Distribution (MODI) Method or uv method
▶ In solving TPs, same Simplex Algorithm is used but in different form
Transportation Problem Solving

▶ Identifying a BFS
▶ North West Corner Method
▶ Least Cost Method
▶ Vogel’s Approximation Method
▶ Finding Optimal Solution
▶ Stepping Stone Method
▶ Modified Distribution (MODI) Method or uv method
▶ In solving TPs, same Simplex Algorithm is used but in different form
▶ In Identifying the BFSs, Vogel’s Approximation method gives the Best BFS
Transportation Problem as an LPP

A firm owns facilites at seven places. It has manufacturing units at A, B and C with daily
outputs 500, 300 and 200 units of an item. It has warehouses at P, Q, R and S with daily
requirements of 180, 150, 350 and 320 units respectively. Per unit shipping charges on
different routes are given below

From/To P Q R S
A 12 10 12 13
B 7 11 8 14
C 6 16 11 7

Formulate the problem as an LPP


Transportation Problem as an LPP, Contd.
From/To P Q R S Supply
A 12 10 12 13 500
B 7 11 8 14 300
C 6 16 11 7 200
Demand 180 150 350 320 1000
Transportation Problem as an LPP, Contd.
From/To P Q R S Supply
A 12 10 12 13 500
B 7 11 8 14 300
C 6 16 11 7 200
Demand 180 150 350 320 1000

Suppose xij be the no. of units to be shipped from Plant i to Warehouse j


Transportation Problem as an LPP, Contd.
From/To P Q R S Supply
A 12 10 12 13 500
B 7 11 8 14 300
C 6 16 11 7 200
Demand 180 150 350 320 1000

Suppose xij be the no. of units to be shipped from Plant i to Warehouse j


Objective function can be written as min Z = 12x11 + 10x12 + 12x13 + 13x14 + . . . + 7x34
Transportation Problem as an LPP, Contd.
From/To P Q R S Supply
A 12 10 12 13 500
B 7 11 8 14 300
C 6 16 11 7 200
Demand 180 150 350 320 1000

Suppose xij be the no. of units to be shipped from Plant i to Warehouse j


Objective function can be written as min Z = 12x11 + 10x12 + 12x13 + 13x14 + . . . + 7x34
Subject to,
Demand Constraints:
Supply Constraints:
x11 + x21 + x31 = 180
x11 + x12 + x13 + x14 = 500
x12 + x22 + x32 = 150
x21 + x22 + x23 + x24 = 300
x13 + x23 + x33 = 350
x31 + x32 + x33 + x34 = 200
x14 + x24 + x34 = 320
North West Corner Rule

Always start with the North West Corner and allocate

From/To P Q R S Supply
12 10 12 13
A 500
7 11 8 14
B 300
6 16 11 7
C 200

Demand 180 150 350 320 1000


North West Corner Rule, Contd.

From/To P Q R S Supply
12 10 12 13
A 180 500 320
7 11 8 14
B 300
6 16 11 7
C 200

Demand 180 150 350 320 1000


0
North West Corner Rule, Contd.

From/To P Q R S Supply
12 10 12 13
A 180 150 500 320 170
7 11 8 14
B 300
6 16 11 7
C 200

Demand 180 150 350 320 1000


0 0
North West Corner Rule, Contd.

From/To P Q R S Supply
12 10 12 13
A 180 150 170 500 320 170 0
7 11 8 14
B 300
6 16 11 7
C 200

Demand 180 150 350 320 1000


0 0 180
North West Corner Rule, Contd.

From/To P Q R S Supply
12 10 12 13
A 180 150 170 500 320 170 0
7 11 8 14
B 300 120
180
6 16 11 7
C 200

Demand 180 150 350 320 1000


0 0 180
0
North West Corner Rule, Contd.

From/To P Q R S Supply
12 10 12 13
A 180 150 170 500 320 170 0
7 11 8 14
B 300 120 0
180 120
6 16 11 7
C 200 200 0

Demand 180 150 350 320 1000


0 0 180
0

Total cost = 12 × 180 + 10 × 150 + 12 × 170 + 8 × 180 + 14 × 120 + 7 × 200 = 10220


Least Cost Method

Look for the Least Cost Cell among all and allocate

From/To P Q R S Supply
12 10 12 13
A 500
7 11 8 14
B 300
6 16 11 7
C 180 200 20

Demand 180 150 350 320 1000


0
Least Cost Method, Contd.

From/To P Q R S Supply
12 10 12 13
A 500
7 11 8 14
B 300
6 16 11 7
C 180 20 200 20 0

Demand 180 150 350 320 1000


0 300
Least Cost Method, Contd.

From/To P Q R S Supply
12 10 12 13
A 500
7 11 8 14
B 300 0
300
6 16 11 7
C 180 20 200 20 0

Demand 180 150 350 320 1000


0 50 300
Least Cost Method, Contd.

From/To P Q R S Supply
12 10 12 13
A 150 500 350
7 11 8 14
B 300 0
300
6 16 11 7
C 180 20 200 20 0

Demand 180 150 350 320 1000


0 0 50 300
Least Cost Method, Contd.

From/To P Q R S Supply
12 10 12 13
A 150 50 500 350 300
7 11 8 14
B 300 0
300
6 16 11 7
C 180 20 200 20 0

Demand 180 150 350 320 1000


0 0 50 300
0
Least Cost Method, Contd.

From/To P Q R S Supply
12 10 12 13
A 150 50 300 500 350 300 0
7 11 8 14
B 300 0
300
6 16 11 7
C 180 20 200 20 0

Demand 180 150 350 320 1000


0 0 50 300
0 0

Total cost = 10 × 150 + 12 × 50 + 13 × 300 + 8 × 300 + 6 × 180 + 7 × 20 = 9620


Vogel’s Approximation Method

In this method, the penalties of not allocating to least cost cell in each row and column
are computed
Select the maximum penalty and the corresponding row or column and allocate in the
least cost cell of that particular row or column

From/To P Q R S Supply I
12 10 12 13
A 500 2
7 11 8 14
B 300 1
6 16 11 7
C 200 200/0 1

Demand 180 150 350 320/120 1000

I 1 1 3 6
Vogel’s Approximation Method, Contd.

From/To P Q R S Supply I II
12 10 12 13
A 500 2 2
7 11 8 14
B 300/120 1 1
180
6 16 11 7
C 200 200/0 1 x

Demand 180/0 150 350 320/120 1000


I 1 1 3 6
II 5 1 4 1
Vogel’s Approximation Method, Contd.

From/To P Q R S Supply I II III


12 10 12 13
A 500 2 2 2
7 11 8 14
B 300/120/0 1 1 3
180 120
6 16 11 7
C 200 200/0 1 x x

Demand 180/0 150 350/230 320/120 1000


I 1 1 3 6
II 5 1 4 1
III x 1 4 1
Vogel’s Approximation Method, Contd.

From/To P Q R S Supply I II III


12 10 12 13
A 150 230 120 500/0 2 2 2
7 11 8 14
B 300/120/0 1 1 3
180 120
6 16 11 7
C 200 200/0 1 x x

Demand 180/0 150/0 350/230/0 320/120/0 1000


I 1 1 3 6
II 5 1 4 1
III x 1 4 1

Total Cost = 10 × 150 + 12 × 230 + 13 × 120 + 7 × 180 + 8 × 120 + 7 × 200 = 9440


Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
▶ All the Non Basic Variables (mn − m − n + 1 in number) cells must be tested
whether allocating at that cell may improve the cost of allocation
Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
▶ All the Non Basic Variables (mn − m − n + 1 in number) cells must be tested
whether allocating at that cell may improve the cost of allocation
▶ Two process of testing namely, Stepping Stone Method and MODI method
Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
▶ All the Non Basic Variables (mn − m − n + 1 in number) cells must be tested
whether allocating at that cell may improve the cost of allocation
▶ Two process of testing namely, Stepping Stone Method and MODI method
▶ Stepping Stone method compares the opportunity cost of allocating to a particular
Non Basic Variable cell
Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
▶ All the Non Basic Variables (mn − m − n + 1 in number) cells must be tested
whether allocating at that cell may improve the cost of allocation
▶ Two process of testing namely, Stepping Stone Method and MODI method
▶ Stepping Stone method compares the opportunity cost of allocating to a particular
Non Basic Variable cell
▶ MODI method is more efficient than Stepping Stone method, where the Duality
concept is utilized to find the best NBV cell to allocate
Testing the Optimality of Initial BFS

▶ Total number of allocations must be at most m + n − 1, representing the Basic


Variables cells
▶ All the Non Basic Variables (mn − m − n + 1 in number) cells must be tested
whether allocating at that cell may improve the cost of allocation
▶ Two process of testing namely, Stepping Stone Method and MODI method
▶ Stepping Stone method compares the opportunity cost of allocating to a particular
Non Basic Variable cell
▶ MODI method is more efficient than Stepping Stone method, where the Duality
concept is utilized to find the best NBV cell to allocate
▶ If number of allocations is less than m + n − 1 then it leads to Degeneracy
MODI Method

▶ For larger problems Stepping Stone method may be tedious


MODI Method

▶ For larger problems Stepping Stone method may be tedious


▶ In MODI method, dual variables ui for rows and vj for columns are assumed
▶ For a particular set of allocations, for all the occupied cells will be assumed to have
cij = ui + vj
MODI Method

▶ For larger problems Stepping Stone method may be tedious


▶ In MODI method, dual variables ui for rows and vj for columns are assumed
▶ For a particular set of allocations, for all the occupied cells will be assumed to have
cij = ui + vj
▶ Calculate the ui and vj for the occupied cells, by starting with u1 = 0
MODI Method

▶ For larger problems Stepping Stone method may be tedious


▶ In MODI method, dual variables ui for rows and vj for columns are assumed
▶ For a particular set of allocations, for all the occupied cells will be assumed to have
cij = ui + vj
▶ Calculate the ui and vj for the occupied cells, by starting with u1 = 0
▶ Calculate the values ∆ij = ui + vj − cij for the unoccupied cells also known as
Opportunity Cost of that unoccupied cell
MODI Method

▶ For larger problems Stepping Stone method may be tedious


▶ In MODI method, dual variables ui for rows and vj for columns are assumed
▶ For a particular set of allocations, for all the occupied cells will be assumed to have
cij = ui + vj
▶ Calculate the ui and vj for the occupied cells, by starting with u1 = 0
▶ Calculate the values ∆ij = ui + vj − cij for the unoccupied cells also known as
Opportunity Cost of that unoccupied cell
▶ If all the Opportunity Costs are negative then the allocation is optimal and unique
otherwise select the cell with highest positive Opportunity Cost provided that cell
makes a loop with other occupied cells in some way
MODI Method

▶ For larger problems Stepping Stone method may be tedious


▶ In MODI method, dual variables ui for rows and vj for columns are assumed
▶ For a particular set of allocations, for all the occupied cells will be assumed to have
cij = ui + vj
▶ Calculate the ui and vj for the occupied cells, by starting with u1 = 0
▶ Calculate the values ∆ij = ui + vj − cij for the unoccupied cells also known as
Opportunity Cost of that unoccupied cell
▶ If all the Opportunity Costs are negative then the allocation is optimal and unique
otherwise select the cell with highest positive Opportunity Cost provided that cell
makes a loop with other occupied cells in some way
▶ If any unoccupied cell has 0 Opportunity Cost with all other as negatives, then
Alternate Optimality exists
MODI Method, Contd.

Start with the Initial BFS obtained by NWC Method and ui , vj are computed

From/To P Q R S Supply ui
12 10 12 13
A 180 150 170 500 0
7 11 8 14
B 300 -4
180 120
6 16 11 7
C 200 200 -11

Demand 180 150 350 320 1000


vj 12 10 12 18

Start with u1 = 0, then v1 + u1 = 12 =⇒ v1 = 12.


MODI Method, Contd.

Start with the Initial BFS obtained by NWC Method and ui , vj are computed

From/To P Q R S Supply ui
12 10 12 13
A 180 150 170 500 0
7 11 8 14
B 300 -4
180 120
6 16 11 7
C 200 200 -11

Demand 180 150 350 320 1000


vj 12 10 12 18

Start with u1 = 0, then v1 + u1 = 12 =⇒ v1 = 12. Similarly, v2 = 10, v3 = 12.


MODI Method, Contd.

Start with the Initial BFS obtained by NWC Method and ui , vj are computed

From/To P Q R S Supply ui
12 10 12 13
A 180 150 170 500 0
7 11 8 14
B 300 -4
180 120
6 16 11 7
C 200 200 -11

Demand 180 150 350 320 1000


vj 12 10 12 18

Start with u1 = 0, then v1 + u1 = 12 =⇒ v1 = 12. Similarly, v2 = 10, v3 = 12. Now,


u2 + v3 = 8 =⇒ u2 = −4 and u2 + v4 = 14 =⇒ v4 = 18.
MODI Method, Contd.

Now compute the opportunity costs for the unoccupied cells using ui + vj − cij

From/To P Q R S Supply ui
12 10 12 13
A 180 150 170 5 500 0
7 11 8 14
B 300 -4
1 -5 180 120
6 16 11 7
C -5 -17 -10 200 200 -11

Demand 180 150 350 320 1000


vj 12 10 12 18
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
▶ Here, both of the unoccupied cells form loops with occupied cells
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
▶ Here, both of the unoccupied cells form loops with occupied cells
▶ Loop for cells {1, 4} and {2, 1}
12 13 12 10 12
170−θ ↓ 5 +θ ↑ 180−θ ↓ 150 170+θ ↑
8 14 7 11 8
180+θ ↑ 120−θ ↓ 1 +θ ↑ -5 180−θ ↓
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
▶ Here, both of the unoccupied cells form loops with occupied cells
▶ Loop for cells {1, 4} and {2, 1}
12 13 12 10 12
170−θ ↓ 5 +θ ↑ 180−θ ↓ 150 170+θ ↑
8 14 7 11 8
180+θ ↑ 120−θ ↓ 1 +θ ↑ -5 180−θ ↓
▶ Always consider the corner cells in constructing any loop
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
▶ Here, both of the unoccupied cells form loops with occupied cells
▶ Loop for cells {1, 4} and {2, 1}
12 13 12 10 12
170−θ ↓ 5 +θ ↑ 180−θ ↓ 150 170+θ ↑
8 14 7 11 8
180+θ ↑ 120−θ ↓ 1 +θ ↑ -5 180−θ ↓
▶ Always consider the corner cells in constructing any loop
▶ What should be the value of θ so that all demand-supply for all the associated
columns and rows are maintained
MODI Method, Contd.
▶ Cells {1, 4} and {2, 1} are having opportunity costs 5 and 1, both positive =⇒
Opportunity to Reduce Cost
▶ Check whether these two unoccupied cells form loop(s) with other occupied cells or
not
▶ Here, both of the unoccupied cells form loops with occupied cells
▶ Loop for cells {1, 4} and {2, 1}
12 13 12 10 12
170−θ ↓ 5 +θ ↑ 180−θ ↓ 150 170+θ ↑
8 14 7 11 8
180+θ ↑ 120−θ ↓ 1 +θ ↑ -5 180−θ ↓
▶ Always consider the corner cells in constructing any loop
▶ What should be the value of θ so that all demand-supply for all the associated
columns and rows are maintained
▶ In case of {1, 4} =⇒ 120 and for cell {2, 1} =⇒ 180
MODI Method, contd.

Out of these two cells, {1, 4} is more promising because of higher opportunity cost
MODI Method, contd.

Out of these two cells, {1, 4} is more promising because of higher opportunity cost For
cell {1, 4}, after re-allocating 120 units, new allocation tableau is

From/To P Q R S Supply ui
12 10 12 13
A 180 150 50 120 500 0
7 11 8 14
B 300 -4
1 -5 300 -5
6 16 11 7
C 0 -12 -5 200 200 -6

Demand 180 150 350 320 1000


vj 12 10 12 13

Here, cell {2, 1} is having positive opportunity cost so, is to be considered for allocation
MODI Method, Contd.

From/To P Q R S Supply ui
12 10 12 13
A -1 150 230 120 500 0
7 11 8 14
B 300 -4
180 -5 120 -5
6 16 11 7
C -1 -12 -5 200 200 -6

Demand 180 150 350 320 1000


vj 11 10 12 13

Here no unoccupied cell is having positive opportunity cost so, this solution is optimal.
Cost is = 12 × 150 + 12 × 230 + 13 × 120 + 7 × 180 + 8 × 120 + 7 × 200 = 9440
This cost is same as the one with VA method
Unbalanced Transportation Problems or/and Prohibited Routes

▶ For unbalanced TPs, Total Availability and Total Demand are not equal i.e.
Pm Pn
i=1 ai ̸= j=1 bj
Unbalanced Transportation Problems or/and Prohibited Routes

▶ For unbalanced TPs, Total Availability and Total Demand are not equal i.e.
Pm Pn
i=1 ai ̸= j=1 bj
▶ If, i=1 ai ≥ nj=1 bj then a Dummy Destination is added and m
Pm P P Pn
i=1 ai ≤ j=1 bj , a
Dummy Source is added
Unbalanced Transportation Problems or/and Prohibited Routes

▶ For unbalanced TPs, Total Availability and Total Demand are not equal i.e.
Pm Pn
i=1 ai ̸= j=1 bj
▶ If, i=1 ai ≥ nj=1 bj then a Dummy Destination is added and m
Pm P P Pn
i=1 ai ≤ j=1 bj , a
Dummy Source is added
▶ As generally, Cost of Transportation is to be minimized so, Cost of Transportations of
these Dummy are kept at 0s
Unbalanced Transportation Problems or/and Prohibited Routes

▶ For unbalanced TPs, Total Availability and Total Demand are not equal i.e.
Pm Pn
i=1 ai ̸= j=1 bj
▶ If, i=1 ai ≥ nj=1 bj then a Dummy Destination is added and m
Pm P P Pn
i=1 ai ≤ j=1 bj , a
Dummy Source is added
▶ As generally, Cost of Transportation is to be minimized so, Cost of Transportations of
these Dummy are kept at 0s
▶ After modification of the original problem, same method is to be followed as in
Balanced ones
Unbalanced Transportation Problems or/and Prohibited Routes

▶ For unbalanced TPs, Total Availability and Total Demand are not equal i.e.
Pm Pn
i=1 ai ̸= j=1 bj
▶ If, i=1 ai ≥ nj=1 bj then a Dummy Destination is added and m
Pm P P Pn
i=1 ai ≤ j=1 bj , a
Dummy Source is added
▶ As generally, Cost of Transportation is to be minimized so, Cost of Transportations of
these Dummy are kept at 0s
▶ After modification of the original problem, same method is to be followed as in
Balanced ones
▶ Some problems may have prohibited routes between Source(s) and Desitnation(s)
where the Transportation costs are assumed as M which is a very big number so that
everytime that(those) particular route(s) will be avoided
Unbalanced Prohibited Routes TP - An Example

Consider the following TP

WH/MKT A B C Supply
1 10 12 7 180
2 14 11 6 100
3 9 5 13 160
4 11 7 9 120
Demand 240 200 220

It is also known that, nothing can be sent from WH 1 to MKT A and WH 3 to MKT C. Find
out the minimum cost of transportation with optimal allocations.
Unbalanced Prohibited Routes TP - An Example
▶ Unbalanced TP with Total Demand = 660 units & Total Supply = 560 units
Unbalanced Prohibited Routes TP - An Example
▶ Unbalanced TP with Total Demand = 660 units & Total Supply = 560 units
▶ So, a Dummy Warehouse will be introduced with 0 transportation costs
Unbalanced Prohibited Routes TP - An Example
▶ Unbalanced TP with Total Demand = 660 units & Total Supply = 560 units
▶ So, a Dummy Warehouse will be introduced with 0 transportation costs
▶ Prohibited routes’ Transportation costs are assumed as M
Unbalanced Prohibited Routes TP - An Example
▶ Unbalanced TP with Total Demand = 660 units & Total Supply = 560 units
▶ So, a Dummy Warehouse will be introduced with 0 transportation costs
▶ Prohibited routes’ Transportation costs are assumed as M
▶ So, Modified problem is

WH/MKT A B C Supply
1 M 12 7 180
2 14 11 6 100
3 9 5 M 160
4 11 7 9 120
5 0 0 0 100
Demand 240 200 220 660
Unbalanced Prohibited Routes TP - An Example
Initial BFS using VAM with cost = 4300

WH/MKT A B C Supply u
M 12 7
1 -1 180 180 15
14 11 6
2 100 14
60 -1 40
9 5 M
3 160 9
0 ↑ 160 ↓
11 7 9
4 120 11
80 ↓ 40 ↑ -6
0 0 0
5 100 -4 -8 100 0

Demand 240 200 220 660


v 0 -4 -8
Unbalanced Prohibited Routes TP - An Example
Different allocation with same cost = 4300

WH/MKT A B C Supply u
M 12 7
1 -1 180 180 15
14 11 6
2 100 14
60 -1 40
9 5 M
3 160 9
80 80
11 7 9
4 120 11
0 0 120 -6
0 0 0
5 100 -4 -8 100 0

Demand 240 200 220 660


v 0 -4 -8
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
▶ Stepping stone and MODI method equires non-zero allocations exactly equal to
m+n−1
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
▶ Stepping stone and MODI method equires non-zero allocations exactly equal to
m+n−1
▶ To remove degenracy, an infinitesimally small quantity ϵ is to be allocated at a cell
which does not form any loop with the occupied cells
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
▶ Stepping stone and MODI method equires non-zero allocations exactly equal to
m+n−1
▶ To remove degenracy, an infinitesimally small quantity ϵ is to be allocated at a cell
which does not form any loop with the occupied cells
▶ Start with the unoccupied cell with smallest cost and check whether it constitutes
any loop with the occupied cells or not
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
▶ Stepping stone and MODI method equires non-zero allocations exactly equal to
m+n−1
▶ To remove degenracy, an infinitesimally small quantity ϵ is to be allocated at a cell
which does not form any loop with the occupied cells
▶ Start with the unoccupied cell with smallest cost and check whether it constitutes
any loop with the occupied cells or not
▶ In case if it requires two or more ϵs then start with smallest cost and then next to
smallest and so on
Degeneracy in Transportation Problems

▶ BFS for a Transpotation problem must contain m + n − 1 non-zero allocations


▶ If a BFS does contain non-zero allocations less than m + n − 1 =⇒ Degeneracy
▶ Stepping stone and MODI method equires non-zero allocations exactly equal to
m+n−1
▶ To remove degenracy, an infinitesimally small quantity ϵ is to be allocated at a cell
which does not form any loop with the occupied cells
▶ Start with the unoccupied cell with smallest cost and check whether it constitutes
any loop with the occupied cells or not
▶ In case if it requires two or more ϵs then start with smallest cost and then next to
smallest and so on
▶ The operations consisting ϵ are k + ϵ = k, k − ϵ = k, 0 + ϵ = ϵ, ϵ + ϵ = ϵ, ϵ − ϵ = 0 and
k×ϵ=0
Degeneracy in Transportation Problems

Determine the optimal solution to the problem given below

To Market
1 2 3 4 Supply
1 6 4 9 1 40
2 20 6 11 3 40
From Plant 3 7 1 0 14 50
4 7 1 12 6 90
Demand 90 30 50 30
Degeneracy in Transportation Problems
Initial Degenerate BFS using VAM with cost = 700

1 2 3 4 Dummy S u

6 4 9 1 0
1 30−θ ↓ -4 -11 10+θ ↑ -2 40 0

20 6 11 3 0
2 40 2
-12 -4 -11 20−θ ↓ 20+θ ↑
7 1 0 14 0
3 50 2
1 +θ ↑ 1 50 -11 ϵ−θ ↓
7 1 12 6 0
4 60 30 -13 -4 -1 90 1

D 90 30 50 30 20 220

v 6 0 -2 1 -2
Degeneracy in Transportation Problems
Revised solution is with same cost = 700

1 2 3 4 Dummy S u

6 4 9 1 0
1 30 -4 -10 10 -2 40 0

20 6 11 3 0
2 40 2
-12 -4 -10 20 20
7 1 0 14 0
3 50 2
ϵ 0 50 -12
7 1 12 6 0
4 60 30 -12 -4 -1 90 1

D 90 30 50 30 20 220

v 6 0 -1 1 -2
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
▶ The resultant matrix known as Opportunity Loss Matrix =⇒ Entries indicate how
much they are away from the highest profit value
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
▶ The resultant matrix known as Opportunity Loss Matrix =⇒ Entries indicate how
much they are away from the highest profit value
▶ The Minimization of Opportunity Loss automatically leads to Maximization of
Profits
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
▶ The resultant matrix known as Opportunity Loss Matrix =⇒ Entries indicate how
much they are away from the highest profit value
▶ The Minimization of Opportunity Loss automatically leads to Maximization of
Profits
▶ The Total Profit of allocation would however be computed based on the original
Profit matrix
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
▶ The resultant matrix known as Opportunity Loss Matrix =⇒ Entries indicate how
much they are away from the highest profit value
▶ The Minimization of Opportunity Loss automatically leads to Maximization of
Profits
▶ The Total Profit of allocation would however be computed based on the original
Profit matrix
▶ For unbalanced problems, a dummy row/column must be added before converting
into Opportunity Loss matrix
Maximization Problem
▶ Though conventionally the TPs are for minimizing the cost of Transportation, the cost
elements can be replaced by unit profit
▶ Convert the problem into an equivalent minimization problem by subtracting each
profit element from the highest profit among them
▶ The resultant matrix known as Opportunity Loss Matrix =⇒ Entries indicate how
much they are away from the highest profit value
▶ The Minimization of Opportunity Loss automatically leads to Maximization of
Profits
▶ The Total Profit of allocation would however be computed based on the original
Profit matrix
▶ For unbalanced problems, a dummy row/column must be added before converting
into Opportunity Loss matrix
▶ For prohibited routes Unit Profit will be replaced by −M before conversion
Maximization Problem

Solve the following TP for maximization of profit

Markets
A B C D S

Warehouses X 12 18 6 25 200
Y 8 7 10 18 500
Z 14 3 11 20 300
D 180 320 100 400 1000
Maximization Problem

▶ Initial BFS x12 = 200, x22 = 100, x24 = 400, x31 = 180, x32 = 20, x33 = 100 =⇒ Profit
= 15, 180
Maximization Problem

▶ Initial BFS x12 = 200, x22 = 100, x24 = 400, x31 = 180, x32 = 20, x33 = 100 =⇒ Profit
= 15, 180
▶ After MODI, x12 = 200, x22 = 120, x24 = 380, x31 = 180, x33 = 100, x34 = 20 =⇒ Profit
= 15, 300
Maximization Problem

▶ Initial BFS x12 = 200, x22 = 100, x24 = 400, x31 = 180, x32 = 20, x33 = 100 =⇒ Profit
= 15, 180
▶ After MODI, x12 = 200, x22 = 120, x24 = 380, x31 = 180, x33 = 100, x34 = 20 =⇒ Profit
= 15, 300
▶ Again, after MODI,
x12 = 200, x22 = 120, x23 = 100, x24 = 280, x31 = 180, x34 = 120 =⇒ Profit = 15, 400
Transshipment Problems

▶ In reality to save cost of transportation, Sources can act Sources for Sources and
Destinations can act as Sources to Destinations =⇒ Interaction between Sources
themselves and Destinations themselves are possible
Transshipment Problems

▶ In reality to save cost of transportation, Sources can act Sources for Sources and
Destinations can act as Sources to Destinations =⇒ Interaction between Sources
themselves and Destinations themselves are possible
▶ Refer the following problem
▶ A firm owns facilites at seven places. It has manufacturing units at A, B and C with
daily outputs 500, 300 and 200 units of an item. It has warehouses at P, Q, R and S
with daily requirements of 180, 150, 350 and 320 units respectively. Per unit shipping
charges on different routes are given below
From/To P Q R S
A 12 10 12 13
B 7 11 8 14
C 6 16 11 7
Transshipment Problems

Along with the above data, some other data are also available

From/To P Q R S
From/To A B C
P 0 6 5 8
A 0 2 8
Q 8 0 3 2
B 5 0 7
R 5 4 0 10
C 10 9 0
S 9 4 8 0
Transshipment Problems

Along with the above data, some other data are also available

From/To P Q R S
From/To A B C
P 0 6 5 8
A 0 2 8
Q 8 0 3 2
B 5 0 7
R 5 4 0 10
C 10 9 0
S 9 4 8 0
In these types of problems, any xii = 0 as no terminal is sending itself
Transshipment Problems

So, the transportation tableau will be an enlarged one

A B C P Q R S Supply
A 0 2 8 12 10 12 13 500
B 5 0 7 7 11 8 14 300
C 10 9 0 6 16 11 7 200
P 12 7 6 0 6 5 8 0
Q 10 11 16 8 0 3 2 0
R 12 8 11 5 4 0 10 0
S 13 14 7 9 4 8 0 0
Demands 0 0 0 180 150 350 320
Transshipment Problems
Transshipment Problems
Transshipment Problems
Transshipment Problems

Suppose for a manufacturing concern, 3 Factories are at X, Y and Z with daily productions
150, 100 and 100 units. The markets where the product is being supplied are at A, B, C
and D with daily demands 80, 100, 70 and 100 units. Between these factories and
markets there are two Warehouses W1 and W2 which dont store anything but sometimes
may cost lower while supplying from Factories to Markets. The unit cost of transportations
from Factories to Warehouses and from Warehouses to Markets are given below.

W1 W2
A B C D
X 20 10
W1 8 9 7 14
Y 15 10
W2 8 7 11 12
Z 10 12

Find out the optimal transportation plan.


Transshipment Problems
▶ The alternative paths are to be evaluated to find out least cost paths between
Factories to Markets via Transshipment nodes i.e. Warehouses
Transshipment Problems
▶ The alternative paths are to be evaluated to find out least cost paths between
Factories to Markets via Transshipment nodes i.e. Warehouses
20 8 10
▶ For example, from X, A can be reached in two ways X−
→ W1 →
− A =⇒ 28 and X −→
8
W2 →− A =⇒ 18
Transshipment Problems
▶ The alternative paths are to be evaluated to find out least cost paths between
Factories to Markets via Transshipment nodes i.e. Warehouses
20 8 10
▶ For example, from X, A can be reached in two ways X− → W1 →− A =⇒ 28 and X −→
8
W2 →− A =⇒ 18
▶ Definitely, the lower cost path will be selected i.e. the second one leading to cost 18
Transshipment Problems
▶ The alternative paths are to be evaluated to find out least cost paths between
Factories to Markets via Transshipment nodes i.e. Warehouses
20 8 10
▶ For example, from X, A can be reached in two ways X− → W1 →− A =⇒ 28 and X −→
8
W2 →− A =⇒ 18
▶ Definitely, the lower cost path will be selected i.e. the second one leading to cost 18

A B C D
X 18 17 21 22 150
Y 18 17 21 22 100
Z 18 19 17 24 100
80 100 70 100 350

VAM will give cost 18 × 50 + 17 × 100 + 22 × 100 + 18 × 30 + 17 × 70 = 6, 530 with degeneracy. Allocate ϵ at
(Y,B) to remove degeneracy.
Check for optimality =⇒ 6, 530
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
▶ Step I: Deduct the smallest element from every element in every row
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
▶ Step I: Deduct the smallest element from every element in every row
▶ Step II: Deduct the smallest element from every element in those column(s) if any
column does not have any zero by Step I
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
▶ Step I: Deduct the smallest element from every element in every row
▶ Step II: Deduct the smallest element from every element in those column(s) if any
column does not have any zero by Step I
▶ Step III: Start assigning to zeros. If any row or column has more than one zero then
only one out of those will be assigned
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
▶ Step I: Deduct the smallest element from every element in every row
▶ Step II: Deduct the smallest element from every element in those column(s) if any
column does not have any zero by Step I
▶ Step III: Start assigning to zeros. If any row or column has more than one zero then
only one out of those will be assigned
▶ Step IV: If its a unique assignment using Step IV, then Stop
Assignment Models
▶ In a Transportation model, if the number of sources and sinks are same with supply at
each source as 1 and demand at each sink also 1, then it is called as Assignment
model
▶ In most general case, objective is to minimize the cost of assignments
▶ To find best possible assignments, the following steps are followed
▶ Aim is to create at least one zero in each row and each column
▶ Step I: Deduct the smallest element from every element in every row
▶ Step II: Deduct the smallest element from every element in those column(s) if any
column does not have any zero by Step I
▶ Step III: Start assigning to zeros. If any row or column has more than one zero then
only one out of those will be assigned
▶ Step IV: If its a unique assignment using Step IV, then Stop
▶ As the number of sources and sinks are same so, each row and each column must
have one assignment so, total number of assignments will be no. of rows or no. of
columns (for balanced case)
Assignment Models
Following is the cost matrix of assigning 6 jobs to 6 machines. Find the optimal assignment.

M1 M2 M3 M4 M5 M6
J1 20 18 21 16 20 19
J2 17 19 20 23 22 21
J3 19 21 18 24 20 26
J4 27 23 24 20 19 22
J5 23 22 20 19 25 24
J6 21 24 25 17 21 18
Assignment Models
Following is the cost matrix of assigning 6 jobs to 6 machines. Find the optimal assignment.

M1 M2 M3 M4 M5 M6
J1 20 18 21 16 20 19
J2 17 19 20 23 22 21
J3 19 21 18 24 20 26
J4 27 23 24 20 19 22
J5 23 22 20 19 25 24
J6 21 24 25 17 21 18

Row minimum subtraction


M1 M2 M3 M4 M5 M6
J1 4 2 5 0 4 3
J2 0 2 3 6 5 4
J3 1 3 0 6 2 8
J4 8 4 5 1 0 3
J5 4 3 1 0 6 5
J6 4 7 8 0 4 1
Assignment Models
Following is the cost matrix of assigning 6 jobs to 6 machines. Find the optimal assignment.

M1 M2 M3 M4 M5 M6
J1 20 18 21 16 20 19
J2 17 19 20 23 22 21
J3 19 21 18 24 20 26
J4 27 23 24 20 19 22
J5 23 22 20 19 25 24
J6 21 24 25 17 21 18

Row minimum subtraction Column minimum subtraction


M1 M2 M3 M4 M5 M6 M1 M2 M3 M4 M5 M6
J1 4 2 5 0 4 3 J1 4 0 5 0 4 2
J2 0 2 3 6 5 4 J2 0 0 3 6 5 3
J3 1 3 0 6 2 8 J3 1 1 0 6 2 7
J4 8 4 5 1 0 3 J4 8 2 5 1 0 2
J5 4 3 1 0 6 5 J5 4 1 1 0 6 4
J6 4 7 8 0 4 1 J6 4 5 8 0 4 0
Assignment Models

If all the rows and columns have at least one zero, start assigning

M1 M2 M3 M4 M5 M6
J1 4 0 5 S
0 4 2
J2 0 S
0 3 6 5 3
J3 1 1 0 6 2 7
J4 8 2 5 1 0 2
J5 4 1 1 0 6 4
J6 4 5 8 S
0 4 0

Cost of Assignment is 109


Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns


Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒


Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
▶ Draw Lines on Unmarked Rows and Marked Columns
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
▶ Draw Lines on Unmarked Rows and Marked Columns
▶ Create some more zeros for possible assignments
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
▶ Draw Lines on Unmarked Rows and Marked Columns
▶ Create some more zeros for possible assignments
▶ Find the minimum element which is uncovered
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
▶ Draw Lines on Unmarked Rows and Marked Columns
▶ Create some more zeros for possible assignments
▶ Find the minimum element which is uncovered
▶ Add that element to all the elements at intersection of lines
Assignment Model - Degeracy

▶ If the number of assignments is lesser than number of rows or columns =⇒ Degeneracy


▶ How to remove Degeneracy? =⇒ Create some more zeros eligible for Assignment
▶ Aim is to draw Minimum Number of Lines which will cover all the Assignments
▶ The Number of Lines must be equal to the Number of Assignments
▶ Mark all the Rows that do not have Assignments
▶ Mark Columns which have zeros in those Marked Rows
▶ Mark Rows that have assignments in the Marked Columns
▶ Continue till no such possibilities are left
▶ Draw Lines on Unmarked Rows and Marked Columns
▶ Create some more zeros for possible assignments
▶ Find the minimum element which is uncovered
▶ Add that element to all the elements at intersection of lines
▶ Deduct that element from the other uncovered elements
Assignment Models - Degeneracy
Find the Optimal Assignments for the following 5 jobs to 5 machines problem

M1 M2 M3 M4 M5
J1 11 6 9 18 11
J2 13 20 6 12 14
J3 5 4 6 6 7
J4 18 9 12 17 15
J5 12 7 15 20 11

Row minimum subtraction


M1 M2 M3 M4 M5
J1 5 0 3 12 5
J2 7 14 0 6 8
J3 1 0 2 2 3
J4 9 0 3 8 6
J5 5 0 8 13 4
Assignment Models - Degeneracy
Find the Optimal Assignments for the following 5 jobs to 5 machines problem

M1 M2 M3 M4 M5
J1 11 6 9 18 11
J2 13 20 6 12 14
J3 5 4 6 6 7
J4 18 9 12 17 15
J5 12 7 15 20 11

Row minimum subtraction Column minimum subtraction


M1 M2 M3 M4 M5 M1 M2 M3 M4 M5
J1 5 0 3 12 5 J1 4 0 3 10 2
J2 7 14 0 6 8 J2 6 14 0 4 5
J3 1 0 2 2 3 J3 0 0 2 0 0
J4 9 0 3 8 6 J4 8 0 3 6 3
J5 5 0 8 13 4 J5 4 0 8 11 1
Assignment Models - Degeneracy

Starting BFS

M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1

Degenerate Solution as number of assignments is 3, must have been 5


Assignment Models - Degeneracy

Mark the rows that don’t have assignments

M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1
Assignment Models - Degeneracy

Mark those columns which have zeros in those marked rows


M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1
Assignment Models - Degeneracy

Mark those rows which have assignmnets in those marked columns

M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1
Assignment Models - Degeneracy

Mark those rows which have assignmnets in those marked columns

M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1

Check whether the new marked rows are having any zeros for which the corresponding
columns are unmarked
Assignment Models - Degeneracy

Mark those rows which have assignmnets in those marked columns

M1 M2 M3 M4 M5
J1 4 0 3 10 2
J2 6 14 0 4 5
J3 0 S
0 2 0
S
 0
S

J4 8 S
0 3 6 3
J5 4 S
0 8 11 1

Check whether the new marked rows are having any zeros for which the corresponding
columns are unmarked
Draw lines on Unmarked Rows and Marked Columns
Assignment Models - Degeneracy

Modified Cost Matrix with more zeros


M1 M2 M3 M4 M5
J1 3 0 2 10 1
J2 6 15 0 4 5
J3 0 1 2 0
S
 S
0
J4 7 S
0 2 5 2
J5 3 S
0 7 10 0

Still Degenerate as 4 Assignments


Assignment Models - Degeneracy

Modified Cost Matrix with more zeros Modified Cost Matrix with more zeros
M1 M2 M3 M4 M5 M1 M2 M3 M4 M5
J1 3 0 2 10 1 J1 2 S
0 1 8 S
0
J2 6 15 0 4 5 J2 6 16 0 4 5
J3 0 1 2 0
S
 S
0 J3 0 2 2 0
S
 S
0
J4 7 S
0 2 5 2 J4 6 0 1 4 1
J5 3 S
0 7 10 0 J5 3 1 7 10 0

Still Degenerate as 4 Assignments Still Degenerate


Assignment Models - Degeneracy

Modified Cost Matrix with more zeros


M1 M2 M3 M4 M5
J1 1 0 1 7 S
0
J2 6 17 0 4 5
J3 0 3 2 0
S
 S
0
J4 5 S
0 S
0 3 S
0
J5 2 1 6 9 0

Still Degenerate as 4 Assignments


Assignment Models - Degeneracy

Modified Cost Matrix with more zeros Modified Cost Matrix with more zeros
M1 M2 M3 M4 M5 M1 M2 M3 M4 M5
J1 1 0 1 7 S
0 J1 2 S
0 1 8 S
0
J2 6 17 0 4 5 J2 6 16 0 4 5
J3 0 3 2 0
S
 S
0 J3 0 2 2 0
S
 S
0
J4 5 S
0 S
0 3 S
0 J4 6 0 1 4 1
J5 2 1 6 9 0 J5 3 1 7 10 0

Still Degenerate as 4 Assignments Still Degenerate 4 Assignments


Assignment Models - Degeneracy

Finally,
M1 M2 M3 M4 M5
J1 0 S
0 S
0 6 S
0
J2 5 17 0 3 4
J3 S
0 4 2 0 S
0
J4 4 0 S
0 3 S
0
J5 1 1 6 8 0

This solution is optimal with cost 43 and Total 5 assignments and each row, each column
has one assignment
Assignment Models - Example

Solve the following Assignment Problem for minimizing the cost of assigning the 4 jobs to
4 machines
M1 M2 M3 M4
J1 3 1 4 4
J2 4 2 5 5
J3 5 3 4 8
J4 4 2 5 9
Assignment Models - Example

Solve the following Assignment Problem for minimizing the cost of assigning the 4 jobs to
4 machines
M1 M2 M3 M4
J1 3 1 4 4
J2 4 2 5 5
J3 5 3 4 8
J4 4 2 5 9

Minimimum cost is 13
Special Cases of Assignment Models

▶ Unbalanced Assignment problems


Special Cases of Assignment Models

▶ Unbalanced Assignment problems


▶ Constrained Assignment problems
Special Cases of Assignment Models

▶ Unbalanced Assignment problems


▶ Constrained Assignment problems
▶ Unique v/s Alternate Optimal Assignment problems
Special Cases of Assignment Models

▶ Unbalanced Assignment problems


▶ Constrained Assignment problems
▶ Unique v/s Alternate Optimal Assignment problems
▶ Maximization problems e.g. Revenue Maximization
Alternate Optimal Assignment problem - Example

Find ou the optimal assignment of Jobs to the Machines

M1 M2 M3 M4 M5
J1 25 18 32 20 21
J2 34 25 21 12 17
J3 20 17 20 32 16
J4 20 28 20 16 27

Optmal cost is 66 and two solutions

J1 → M2 J1 → M2
J2 → M4 J2 → M4
J3 → M5 J3 → M5
J4 → M1 J4 → M3
Revenue Maximization

A Company wants to employ Sales Representatives to Districts where it operates.


Estimates of Sales Revenue in thousands of Rupees for each SR in differennt districts are
given in the following table. Find out the optimal employment of SR to the Districts to
maximize the Revenue.
D1 D2 D3 D4 D5
S1 40 46 48 36 48
S2 48 32 36 29 44
S3 49 35 41 38 45
S4 30 46 49 44 44
S5 37 41 48 43 47
Revenue Maximization
Find out the Opportunity Loss Matrix
D1 D2 D3 D4 D5
S1 9 13 1 13 1
S2 1 17 13 20 5
S3 0 14 8 11 4
S4 19 3 0 5 5
S5 12 8 1 6 2

Total 4 Optimal Assignments are possible each leading to Revenue Rs 2,31,000


Assignments are:
S1→D2, S2→D1, S3→D5, S4→D3, S5→D4
S1→D2, S2→D5, S3→D1, S4→D3, S5→D4
S1→D2, S2→D5, S3→D1, S4→D4, S5→D3
S1→D2, S2→D1, S3→D5, S4→D4, S5→D3

You might also like