Original article | doi: 10.25007/ajnu.v6n3a78
46 Academic Journal of Nawroz University (AJNU)
Using Swarm Intelligence for solving NP-
Hard Problems
Saman M. Almufti
Department of Computer Science and Information Technology, College of Computer Science & Information
Technology, Nawroz University, Duhok, Iraq
ABSTRACT
Swarm Intelligence algorithms are computational intelligence algorithms inspired from the collective behavior of
real swarms such as ant colony, fish school, bee colony, bat swarm, and other swarms in the nature. Swarm
Intelligence algorithms are used to obtain the optimal solution for NP-Hard problems that are strongly believed
that their optimal solution cannot be found in an optimal bounded time. Travels Salesman Problem (TSP) is an
NP-Hard problem in which a salesman wants to visit all cities and return to the start city in an optimal time. This
article applies most efficient heuristic based Swarm Intelligence algorithms which are Particle Swarm
Optimization (PSO), Artificial Bee Colony (ABC), Bat algorithm (BA), and Ant Colony Optimization (ACO)
algorithm to find a best solution for TSP which is one of the most well-known NP-Hard problems in
computational optimization. Results are given for different TSP problems comparing the best tours founds by BA,
ABC, PSO and ACO.
KEY WORDS: Swarm Intelligence, NP-Hard problem, Particle Swarm Optimization (PSO), Artificial Bee Colony
(ABC), Bat algorithm (BA), and Ant Colony Algorithm (ACO)
1. INTRODUCTION
Swarm Intelligence algorithms are computational
intelligence techniques studies the collective behavior
in decentralized systems (Almufti, 2015). Such systems
are made up of a population of simple individual’s
agents interacting locally with each other and with the
environment around themselves (Almufti, 2015). This
paper focuses on the comparative analysis of most
successful methods of optimization techniques inspired
by Swarm Intelligence (SI): Ant Colony Optimization
(ACO) and Particle Swarm Optimization (PSO), Bat
Colony Optimization (BA), and Artificial Bee Colony
(ABC) for solving one of the most well-known NP-
Hard problems which is called Travels Salesman
Problem (TSP) (Almufti, 2015) , (Andrej Kazakov,
2009).
2. OVERVIEW OF SWARM INTELLIGENCE
ALGORITHMS
Swarm intelligence(SI),which is an artificial intelligence
(AI) field, is concerned with the designing of intelligent
interactive multi-agent systems that cooperate to gather
to achieve a specific goal. Swarm intelligence is defined
by Dorigo M as “The emergent collective intelligence of
groups of simple agents”( Li, Y.: 2010). Swarm-based
algorithms are inspired from behaviors of some social
living beings (insects, animal, and bacteria’s) in the
nature, such as ants, birds, bats, bees, termites, and
fishes. The most remarkable features of swarm systems
are Self-organization and decentralized control that
naturally leads to an emergent behavior in the colony.
Emergent behavior is an interactive property that
emerges a local interaction among all system
components (agents) and it is not possible to be achieved
alone by any agent in the system (Almufti, 2015). In
computer science there are many algorithms that are
designed as an inspiration of real collective behavior
systems in the nature, swarm intelligence algorithms
includes Ant Colony Optimization (ACO), Particle
Swarm Optimization (PSO), Artificial Bee Colony (ABC),
Artificial Immune System, Bat algorithm, Bacterial
Foraging, Stochastic diffusion search, Glowworm Swarm
Optimization, Gravitational search algorithm, Cat
Swarm Optimization, and other optimization algorithms
(Almufti, 2015).
Swarm intelligence works on two basic principles: self-
Academic Journal of Nawroz University
(AJNU) Volume 6, No 3(2017), 10 pages
Received 1 May 2017; Accepted 1 August 2017
Regular research paper: Published 8 August 2017
Corresponding author’s e-mail: Saman.almofty@gmail.com
Copyright ©2017 Saman M. Abdulrahman.
This is an open access article distributed under the Creative
Commons Attribution License.