Anda di halaman 1dari 15

MILLER TUCKER ZEMLIN

(MTZ) FORMULATION OF
TRAVELLING SALESMAN
PROBLEM (TSP)
Group 5:

Atibhi Lehakra (PGP/18/131)

Ayush Goel (PGP/18/132)


Darshan J (Pgp/18/135)

Nikita Verma (PGP/18/150)

TSP DEFINITION
The Traveling Salesman Problems (TSP) is stated as follows:
given a finite number of cities along with the cost or the distance
of travel between each pair of them, find the cheapest or shortest
way of visiting all the cities and returning to the starting point

ASSUMPTIONS
Symmetric

Costs: We assumed that the cost for travelling from city


A to city B is the same as that of travelling from city B to city A.
Connected graph: There is a connection or path from any node to
any other node in the graph
Triangle inequality: cij + cjk cik i, j, k N, the direct connection
between two nodes is always the cheapest
Closed tour: The salesman has to return to the node where he starts
the tour; this node is usually referred to as the base city or depot.
This assumption is called a closed tour. For a closed tour any node
can be selected as the starting node

TRAVELLING SALESMAN PROBLEM


Minimize Z

C
i 1

j 1

ij

X ij

Subject to
n

j 1
n

i 1

X ij

X ij 1,

i 1, 2, ...n

X ij 1,

j 1, 2, ...n

1, if travelled from city i to city j



0, otherwise

Cij

Cost of traveling from city i to city j ( or )

Distance from cit y i to city j (or)


Travel time between city i to city j

Elimination of subtours
Miller-Tucker Zemlin (MTZ) Formulation (1960)
ui - uj n(1 xij)
for i = 2, , n, i j, j = 2,,n

ADDITIONAL CONSTRAINTS

TSP assumes that:


We can start from any city
We can travel from any city to every other city
We have to cover each and every city in the loop and return back to the starting city

Dharmadam Islands
Mahe

Hotel Sangamam
Downtown
Taj Gateway
Kingsbay
Beach Hotel

Chill Out
Milma, IIM Kozhikode
ITC Kadavu
Mezban

USING TSP SOLVER


TSP Solver screenshot The matrix containing all the prices in INR that we
found for our travel.
Milma

Taj

Kadavu

Dharmadam

Mahe

Kingsbay

Down Town

Mezban

Sangamam

Chillout

Beach Hotel

145

194

834

695

154

140

143

10

2.5

157

Taj

1450

173

740

606

5.5

7.5

32

196

140

12

Kadavu

194

173

861

727

147

168

123

203

211

150

Dharmadam

834

740

861

142

739

740

759

556

827

745

Mahe

695

606

727

142

605

600

626

428

693

611

Kingsbay

154

5.5

147

739

605

12

34

152

144

Down Town

140

7.5

168

740

600

12

30

190

135

16

Mezban

143

32

123

759

626

34

30

163

156

27

Sangamam

10

196

203

556

428

152

190

163

7.5

162

Chillout

2.5

140

211

827

693

144

135

156

7.5

155

Beach Hotel

157

12

150

745

611

16

27

162

155

Location
Milma

SOLUTION

S. No.

Cost

City

194

Kadavu

123

Mezban

11

27

Beach Hotel

Kingsbay

5.5

Taj

7.5

Down Town

600

Mahe

142

Dharmadam

556

Sangamam

10

7.5

Chillout

2.5

Milma

Total

1673

Dharmadam Islands
Mahe

Hotel Sangamam
Downtown
Taj Gateway
Kingsbay
Beach Hotel

Chill Out
Milma, IIM Kozhikode
ITC Kadavu
Mezban

Beach Hotel
Kings Bay
Taj
Downtown
Kadavu
Mezban

Mahe

Milma

Chill Out
Sangamam

Dharmadam

PROBLEMS
Fewer constraints, but harder to solve. The LP relaxation is not as tight. Okay for small
problems, but is bad for large ones
Base Node : Feasibilityandfinalsolutionmaydependonwhichnodeisselectedasthe
basecity.
Non Complete Graphs:
TheMTZSECsmayfailtofindafeasiblesolutionevenifsuchexists.
Problemswithfeasibilitymayoccurinnoncompletegraphs,whereall
feasiblesolutionsrequiresomenodestobevisited twice.
Global Optimal Solution: Problem with global optimal solution occurs when:
Globaloptimalsolution
mayoccurincompletegraphswherethetriangleinequalitydoesnotapply,
Global optimalsolutionrequiressomenodestobevisitedmorethanonce
Itisthereforeimportanttobe
awareofthesetwosituationswhereapplyingtheMTZSECsmaymakethefinalsolutionsensiti
veasto
whichnodeisselectedasthestartingnode.Theywillneverfailiftheglobaloptimalsolutionvisi
tseachnodeonlyonce.

ADVANTAGES

Small size :

we need only n extra variables and roughly n^2/2 extra constraints

Changeable preferences:

if it is preferable to visit, say, city i early in the tour, one can easily model this by
adding a term ui with some >0 to the objective.

THANK
YOU!!!

Anda mungkin juga menyukai