Scribd
Upload
39 views

Traveling Salesman Problem (EXT.) : Prof. U. K. Bhattacharya

This document discusses different approaches to solving the traveling salesman problem (TSP), including heuristic approaches like nearest neighbor and savings algorithms. It also covers variations of the TSP, such as multiple TSP, time-constrained TSP, and pick-up and delivery TSP. Examples are provided to illustrate the savings algorithm approach. The document concludes by noting that TSP and its variations have applications in scheduling and logistics optimization problems.

Uploaded by

JAYANT PATODIA PGP 2019-21 Batch
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
39 views

Traveling Salesman Problem (EXT.) : Prof. U. K. Bhattacharya

This document discusses different approaches to solving the traveling salesman problem (TSP), including heuristic approaches like nearest neighbor and savings algorithms. It also covers variations of the TSP, such as multiple TSP, time-constrained TSP, and pick-up and delivery TSP. Examples are provided to illustrate the savings algorithm approach. The document concludes by noting that TSP and its variations have applications in scheduling and logistics optimization problems.

Uploaded by

JAYANT PATODIA PGP 2019-21 Batch
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 18

TRAVELING

SALESMAN
PROBLEM
[EXT.]

PROF. U. K.
BHATTACHARYA
Heuristic Approach.

Due to the difficulty of the TSP, Many Heuristics (Approximate) procedures have been
developed these heuristic may be compared analytically. Computational experimentation
may be used to compare the performance of these heuristic.
Classification.
Heuristic procedures can be classified into three broad classes (i) tour construction
procedure (ii) tour improvement procedure and (iii) composite procedures .
Nearest neighbor procedure.
Step 1. Start with any node as the beginning of a path.

Step 2. Find the node closest to the last node added to the path. Add this node
to the path.

Step 3. Repeat step 2 until all nodes are contained in the path. Then, join the
first and last nodes.
Application Exercise Q1
Consider the network as given in picture 3. Find the minimal distance
Hamiltonian circuit.

5 9

17
1

15
10
6
7

8
Clark and Wright saving approach.
Procedure

Step 1. Select any nodes as the central depot which we denote as node 1.

Step 2. Compute savings sij  c1i  cij  cij for I, j =1,3,……..n.

Step 3. Order the saving from largest to smallest.

Step4. Starting at the top of the savings list and moving downwards, from larger
sub tours by linking appropriate nodes I and j. Repeat until a tour is formed.
Solution:
Step 1: Node 1 is the central depot.
Savings Matrix
Step 2: Savings matrix. Here the network is a symmetric network.

Savings for 1 2 3 4 5
arc (I,j)

1 --- --- --- --- ---

2 --- --- 13 0 ---

3 --- 13 --- 20 ---

4 --- 0 20 --- 8

5 --- --- --- 8 ---


Step 3:

Savings for arc (i.j) in sorted ( decreasing order) order.

Arc Savings

3-4 & 4-3 20

2-3 & 3-2 13

4-5 & 5-4 8

2-4 & 4-2 0


Step-4
Step 4: Link the appropriate arcs to form a tour.

1-2-3-4-5-1 is the required Hamiltonian cycle with a distance of 35.


Example 2 [Contd.]
Savings for arc (i.j) in sorted ( decreasing order) order.
ecreasing order) order.

Savings in sorted order.


Arc Savings

3-6 & 6-3 20

3-4 & 4-3 20

4-6 &6-4 15

2-3 & 3-2 13

4-5 & 5-4 8

2-4 & 4-2 0


Final Solution:
Route is : 1-2-3-6-4-5-1.

Distance : 48 Units.
Nearest insertion approach.

Procedure:

Step 1. Start with a sub graph consisting node I only.

Step 2. Find node k such that cik is minimal and form the sub tour i-k-i.

Step 3. Selection step: Given a sub tour, find node k not in the sub tour closest to any

node in the sub tour.

Step 4. Insertion step: Find the arc (I,j) in the sub tour which minimizes cik  c kj  cij .

Insert k between I and j.

Step 5. Go to step 3 unless we a Hamiltonian cycle.


Savings Approach for VRP
References:

1. Clark, C. and Wright, J. Scheduling of Vehicles from central depot to a number of


delivery points. Operations Research, Vol. 12, 1964, PP-568-581.

2 A computerized approach to the New York city school bus routing problem, Jeffrey
Braca, ,Julien Bramel, Bruce Posner, and David Simchi- Levy, IIE Transactions, Vol 29,
1997,PP. 693-702.

3. Vehicle Routing in Large Organizations: A Case Study, C. Sateesh, Pradip K. Ray,


IJOPM, PP 71-78.
Variations in Traveling Salesman Problem
A. Multiple Traveling Salesman Problems

The multiple traveling salesman problem is a generalization of the TSP where


there is a need to account for more than one salesman or vehicle. Multiple TSP
arise in various scheduling and sequencing applications. It has already been
proved to be an appropriate model for the problem of bank messenger
scheduling, where a crew of messengers picks up deposits at branch banks and
returns them to the central office for processing.
B. Time Constrained Traveling Salesman Problem.

Suppose that each ordered pain (i, j) of nodes has associated with it a net profit
and a travel time. In the time constrained TSP, the objective is to find a sub tour
that begins and ends at the origin (node 1), which maximized profit while
requiring less than τ units of time. This is a much more realistic statement a
traveling Salesman faces and it has been the focus of research attention by,
Gensch (1978), Golden et al (1981).
C.Pick-up and Delivery Traveling Salesman Problem.
The Traveling Salesman Problems with pickup and delivery (TSPPD) is an
extension of the traditional Traveling Salesman Problem (TSP). In the PDTSP
problem however the customers might require both pickup and deliveries
where the origin and destination of the demands are at the same depot. The
vehicle has limited capacity and could possibly be overloaded if the amount of
the pickups is bigger than the free space of the vehicle. The objective function
of the problem is still to find the shortest tour which visits all the customers
while not violating the capacity constraints. One typical example of this
problem is breweries which delivers full bottles of beer and mineral water to
its customers and collect empty bottles on the same tour.

You might also like