EUROPEAN ACADEMIC RESEARCH

Vol. IV, Issue 3/ June 2016

Impact Factor: 3.4546 (UIF)

ISSN 2286-4822 DRJI Value: 5.9 (B+)
www.euacademic.org

Solving the Traveling Salesman Problem using

Google Services and Simulated Annealing
Algorithm

LAURIK HELSHANI1
PhD(c), European University of Tirana
Tirana, Albania

Abstract:
The concept of the traveling salesman problem is to seeking a
tour of a specified number of cities (visiting each city exactly once and
returning to the starting point) where the length of the tour is
minimized. The mathematical modeling of this problem has to do with
the theory of graphs.
In this paper, I examine simulated annealing algorithm in the
context of solving the traveling salesman problem. Test-tool
programming is done in JAVA programming language and his work is
displayed graphically using the 2D library.
The whole system is programmed as client-server system using
RESTFul web services, Google-services and Simulated Annealing
Algorithm.
Google services are used to determine the optimum route on
Google map and solve the Travelling Salesman problem.

Key words: Traveling Salesman Problem (TSP); Simulated

Annealing Algorithem (SA); Google Services; RESTful; JAVA

1
Laurik Helshani is PhD candidate at Faculty of Economics. Profile:
Management Information Systems in European University of Tirana (UET).
His PhD thesis: ―Solving Travelling Salesman Problem using Genetic
Algorithm and programming android smart-phones‖.

2321
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

I. INTRODUCTION

The Travelling Salesman Problem is an optimization problem

which has various applications such as: vehicle routing and
scheduling, planning and logistics.
The Travelling Salesman Problem (TSP) is a classic
combinatorial optimization problem, which is simple to state
but very difficult to solve. This problem is known to be NP-
hard, and cannot be solved exactly in polynomial time. [1]
This paper proposes an effective local search algorithm
based on simulated annealing technique to solve the TSP.
Google map is used to provide the user an intuitive and
highly responsive mapping interface, to help him by finding and
marking the places he wants to visit. My web-based application
and the results are displayed on Google maps.

II. RELATED WORK

Goossaert, in the article [8] presented simulated annealing, an

optimization technique aimed to find an approximation to the
global minimum of a function. He explained how this technique
applies to the traveling salesman problem. After some tests, he
managed to find a good set of parameters for the algorithm
which is adapted to solve the problem. Finally, he applied the
algorithm and he found an optimal path in a 50-city tour. The
data he is using are GPS coordinates of 50 European cities. A
32% improvement is observed from the initial tour to the
optimal tour. The author provided an implementation in
Python, along with graphic visualization of the solutions.
In the article [9], Arostegui, Kadipasaoglu &
Khumawala compared to the relative performance of Tabu
Search (TS), Simulated Annealing (SA) and Genetic Algorithms
(GA) on various types of Facility Location Problems (FLP)
under time-limited, solution-limited, and unrestricted
conditions. The results indicate that TS shows very good

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2322
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

performance in most cases. It was concluded that the

performance of TS, SA and GA for the different types of FLP is
situational principally based on the measure of comparability.

III. MATHEMATICAL FORMULATIONS OF SHORTEST

TSP

The TSP can be defined on a complete undirected graph G = (V,

E) if it is symmetric or on a directed graph G = (V, A) if it is
asymmetric. The set V ={1, . . . , n} is the vertex set, E = {(i, j): i,
j ∈V, i < j} is an edge set and A = {(i, j ): i, j ∈V, i ≠ j} is an arc
set. A cost matrix C = (cij ) is defined on E or on A. The cost
matrix satisfies the triangle inequality whenever cij ≤ cik + ckj ,
for all i, j, k. In particular, this is the case of planar problems
for which the vertices are points Pi = (Xi, Yi ) in the plane, and

cij = √ (1)

is the Euclidean distance. The triangle inequality is also

satisfied if cij is the length of a shortest path from i to j on G.
[2]

A. Graph theorical formulation

Find the shortest Hamiltonian circuit in a complete graph
where the nodes represent cities. The weights on the edges
represent the distance between cities. The cost of the tour is the
total distance covered in traversing all cities. [3]

B. Cost function
If the problem consists of n cities ci, i = 1 … n, any tour can be
represented as a permutation of numbers 1 to n. [3]
d(ci, cj) = d(cj, ci) (2)

is the distance between ci and cj.

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2323
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

Given a permutation π of the n cities, v i and vi+1 are adjacent

cities in the permutation. The permutation π has to be found
that minimizes [3]:
∑ (3)

The size of the solution space is (n-1)!/2 [3]

IV. GOOGLE SERVICES

You have to add Map API key to your application to access the
Google maps server with the maps API. Register in the Google
APIs Console to get a Map API key.

A. Google maps
Google maps provide an intuitive and highly responsive
mapping interface with aerial imagery and detailed street data.
In addition, map controls and overlays can be added to the map
so that users can have full control over map navigation. Map
panning can also be performed by dragging the map via the
mouse or by using ―arrow‖ keys on a keyboard. Google maps can
be customized according to application specific needs. [4].

B. Geo- coding
GeoCoder is a class for handling geo-coding and reverse geo-
coding. Geo-coding is a process of converting addresses into
geographical coordinates (latitudes and longitudes). Reverse
geo-coding is the process of transforming (latitude, longitude)
into addresses. [4]

C. Over layers
Overlays are objects on the map that are bound to
latitude/longitude coordinates. Google Maps has several types
of overlays:
1) Marker - Single locations on a map. They are used to
identify locations on the map.

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2324
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

2) Polyline - Series of straight lines on a map.

D. Directions
The Google Maps Directions API is a service that calculates
directions between locations using an HTTP request.

E. Google Places
The Google Places API Web Service is a service that returns
information about places — defined within this API as
establishments, geographic locations, or prominent points of
interest — using HTTP requests. Place Searches return a list of
places based on a user's location or search string.

V. SIMULATED ANNEALING ALGORITHM (SA)

A. Strategy
The information processing objective of the technique is to
locate the minimum cost configuration in the search space. The
algorithms plan of action is to probabilistically re-sample the
problem space where the acceptance of new samples into the
currently held sample is managed by a probabilistic function
that becomes more discerning of the cost of samples it accepts
over the execution time of the algorithm. [5]

B. Working of SA
In simulated annealing algorithm, the problem of finding an
optimal solution is analogous to a process of reducing the
overall ―energy" of a system via state transitions. In the
traveling salesman problem, a state is defined as a route, or a
permutation of the cities to be visited. The energy of a state is
the total distance of a route. In every iteration of the algorithm,
a candidate state transition, also known as a neighbor of the
current state, is generated by exchanging. The neighbor route is
then evaluated by an acceptance function that makes the state
transition if the neighbor route has a lower energy or chosen

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2325
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

with a probability that depends on the difference between the

energies and on a global parameter T (called the temperature).
[6]

C. Testing and result interpretation

Test-tool programming is done in JAVA programming language
and his work is displayed graphically using the 2D library of
JAVA.
The tool simulates graphically, what the algorithm is
doing, generating random points in the XY-plane that
symbolize places to visit, and draws straight lines between
points that symbolize the roads that lead to these places.
The algorithm is tested on computer with these
characteristics:
Operation System: Windows 7 Ultimate, 32-bit
CPU: Intel Core 2 Duo, 1.83 GHz
RAM: 4 GB

The
initial
Cooling The length of the route Runtime
#places length Optimization in %
rate after optimization in sec.
of the
route
30 1*e-4 14.9 4.3 345.360 0.04
30 1*e-5 15.2 4.7 323.509 0.335
50 1*e-5 26.1 6.4 405.615 0.441
50 1*e-6 28.1 6.1 459.755 2.974
100 1*e-4 52.5 9.9 530.889 0.167
100 1*e-5 56.3 8.2 690.034 0.656
150 1*e-4 83.9 15.4 543.696 0.069
150 1*e-5 82.2 11.0 747.626 0.915
150 1*e-6 82.6 10.2 808.325 6.635
150 1*e-7 80.1 10.5 763.005 68.051
Tastcases oF SA- Algorithm

As seen from the above table, there are two key variables,
which effects directly to the performance of the simulated
annealing algorithm work, to its termination and to the final
result (optimization of the route of travelling salesman). These
variables are: Number of places to visit (#places) and
temperature step (The degree of cooling temperature). The
EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016
2326
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

other four columns: The initial length of the route, the length of
the route after optimization, Optimization expressed as a
percentage and runtime in seconds are results of the tool.
Optimization reaches the culmination (~808.325%), if
the temperature decreases with 1*e-6 cooling rate and it
happens in 6.635 seconds.

Fig 1. Programmed Test-tool in JAVA

VI. THE PROGRAMED SYSTEM

The whole system consists from web application as a client and

a system as a server.

A. Client side
The system is implemented to solve the Travelling Salesman
problem to determine the optimum route on Google map using

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2327
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

Google API and Simulated Annealing Algorithm in web –

application. The system starts from getting different Geo
Locations (Latitude and Longitude) on the map triggered by the
user. The user clicks on the Google Map and markers the place
he wants to visit.

Fig 2. User marks the places on Google map through markes

Once the user has chosen places to visit and has select function
"Optimize and map the route", the system derives all the
necessary information from Google Map and packs them into
JSON. JSON data are sent by web application to the server via
HTTP-GET request through JQuery.

Fig 3. The optimized route on Google map drawn through polilynes

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2328
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

B. Server side
Travelling Salesman Problem has been analyzed and Simulated
Annealing has been implemented, in Java, as RESTful web
service, successfully. REST involves the transfer of
resources between clients and servers. This server side system
consumes JSON formatted data, produces XML- formatted data
and is running on tomcat sever.
The server consists of the following classes:
a) Place models a google place. It contains all the
characteristics of a google place: id, name, address,
latitude and longitude
b) Tour  stores a candidate tour and is a list of places
(ListArray<Place>)
c) TourManager  holds the places of a tour and is a list of
places (ListArray<Place>)
d) SimulatedAnnealing

VII. CONCLUSIONS

Simulated Annealing Algorithm is based on neighborhood

search. It has a strong analogy to the simulation of cooling of
material.
Simulated Annealing Algorithm has a great influence on
the optimization of round trip of TSP, based on results in above
table.
Simulated Annealing reaches an approximate solution
(close to the optimal).
The purpose of this paper is to show how to use Google -
Services in combination with Simulated Annealing algorithm to
solve TSP problem. This purpose makes working to discern
from the papers of the earlier or existing.

ACKNOWLEGMENT
This paper is made possible through the help and support from my family.

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2329
Laurik Helshani- Solving the Traveling Salesman Problem using Google
Services and Simulated Annealing Algorithm

REFERENCES

1. J.-Y. Potvin, ―The traveling salesman problem‖, Annals

of Operations Research 63, Canada, 2006, p.: 339-370.
2. R. Matai, S. Singh, M. Mittal: ―Traveling Salesman
Problem: an Overview of Applications, Formulations,
and Solution Approaches, Traveling Salesman Problem‖,
Theory and Applications, India, 2010.
3. P. Eles: ―Heuristic Algorithms for Combinatorial
Optimization Problems: Simulated Annealing‖,
Department of Computer and Information Science,
Linköpings Universitet, Sweden, 2010.
4. Anupriya, Mansi Saxena: An Android Application for
Google Map Navigation System Implementing
Travelling Salesman Problem, in Proceedings of
International Journal of Computer & Organization
Trends –Volume3 Issue4, 2013.
5. F. Yang: ―Solving Traveling Salesman Problem Using
Parallel Genetic Algorithm and Simulated Annealing‖,
2010.
6. M. Kämpf: ―Probleme der Tourenbildung‖, Technische
Universität Chemnitz, Deutschland, 2006.
7. J. Brownlee: ―Clever Algorithms Nature-Inspired
Programming Recipes‖, Revision 2. 2012, ISBN: 978-1-
4467-8506-5; p 175-176
8. D. Goossaert. (2010, April 6). Simulated annealing
applied to the traveling salesman problem. Available at:
http://codecapsule.com/2010/04/06/simulated-annealing-
traveling-salesman/
9. M. Arostegui, S. Kadipasaoglu, B. Khumawala, ―An
empirical comparison of Tabu Search, Simulated
Annealing, and Genetic Algorithms for facilities
location‖, International Journal of Production
Economics, vol 103, no 2, (2006)

EUROPEAN ACADEMIC RESEARCH - Vol. IV, Issue 3 / June 2016

2330