The Ant Lion Optimizer
Seyedali Mirjalili
⇑
School of Information and Communication Technology, Griffith University, Nathan, Brisbane, QLD 4111, Australia
Queensland Institute of Business and Technology, Mt Gravatt, Brisbane, QLD 4122, Australia
article info
Article history:
Received 14 December 2014
Received in revised form 9 January 2015
Accepted 13 January 2015
Available online 3 March 2015
Keywords:
Optimization
Benchmark
Constrained optimization
Particle swarm optimization
Algorithm
Heuristic algorithm
Genetic algorithm
abstract
This paper proposes a novel nature-inspired algorithm called Ant Lion Optimizer (ALO). The ALO
algorithm mimics the hunting mechanism of antlions in nature. Five main steps of hunting prey such
as the random walk of ants, building traps, entrapment of ants in traps, catching preys, and re-building
traps are implemented. The proposed algorithm is benchmarked in three phases. Firstly, a set of 19
mathematical functions is employed to test different characteristics of ALO. Secondly, three classical
engineering problems (three-bar truss design, cantilever beam design, and gear train design) are solved
by ALO. Finally, the shapes of two ship propellers are optimized by ALO as challenging constrained real
problems. In the first two test phases, the ALO algorithm is compared with a variety of algorithms in the
literature. The results of the test functions prove that the proposed algorithm is able to provide very com-
petitive results in terms of improved exploration, local optima avoidance, exploitation, and convergence.
The ALO algorithm also finds superior optimal designs for the majority of classical engineering problems
employed, showing that this algorithm has merits in solving constrained problems with diverse search
spaces. The optimal shapes obtained for the ship propellers demonstrate the applicability of the proposed
algorithm in solving real problems with unknown search spaces as well. Note that the source codes of the
proposed ALO algorithm are publicly available at http://www.alimirjalili.com/ALO.html.
Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction
In recent years metaheuristic algorithms have been used as pri-
mary techniques for obtaining the optimal solutions of real engi-
neering design optimization problems [1–3]. Such algorithms
mostly benefit from stochastic operators [4] that make them dis-
tinct from deterministic approaches. A deterministic algorithm
[5–7] reliably determines the same answer for a given problem
with a similar initial starting point. However, this behaviour results
in local optima entrapment, which can be considered as a disad-
vantage for deterministic optimization techniques [8]. Local
optima stagnation refers to the entrapment of an algorithm in local
solutions and consequently failure in finding the true global opti-
mum. Since real problems have extremely large numbers of local
solutions, deterministic algorithms lose their reliability in finding
the global optimum.
Stochastic optimization (metaheuristic) algorithms [9] refer to
the family of algorithms with stochastic operators including evolu-
tionary algorithms [10]. Randomness is the main characteristic of
stochastic algorithms [11]. This means that they utilize random
operators when seeking for global optima in search spaces.
Although the randomised nature of such techniques might make
them unreliable in obtaining a similar solution in each run, they
are able to avoid local solutions much easier than deterministic
algorithms. The stochastic behaviour also results in obtaining dif-
ferent solutions for a given problem in each run [12].
Evolutionary algorithms search for the global optimum in a
search space by creating one or more random solutions for a given
problem [13]. This set is called the set of candidate solutions. The
set of candidates is then improved iteratively until the satisfaction
of a terminating condition. The improvement can be considered as
finding a more accurate approximation of the global optimum than
the initial random guesses. This mechanism brings evolutionary
algorithms several intrinsic advantages: problem independency,
derivation independency, local optima avoidance, and simplicity.
Problem and derivation independencies originate from the con-
sideration of problems as a black box. Evolutionary algorithms only
utilize the problem formulation for evaluating the set of candidate
solutions. The main process of optimization is done completely inde-
pendent from the problem and based on the provided inputs and
received outputs. Therefore, the nature of the problem is not a con-
cern, yet the representation is the key step when utilizing evolution-
ary algorithms. This is the same reason why evolutionary algorithms
do not need to derivate the problem for obtaining its global optimum.
http://dx.doi.org/10.1016/j.advengsoft.2015.01.010
0965-9978/Ó 2015 Elsevier Ltd. All rights reserved.
⇑
Address: School of Information and Communication Technology, Griffith
University, Nathan, Brisbane, QLD 4111, Australia.
E-mail address: seyedali.mirjalili@griffithuni.edu.au
URL: http://www.alimirjalili.com
Advances in Engineering Software 83 (2015) 80–98
Contents lists available at ScienceDirect
Advances in Engineering Software
journal homepage: www.elsevier.com/locate/advengsoft