Archives of Computational Methods in Engineering (2021) 28:1397–1416

https://doi.org/10.1007/s11831-020-09420-6

ORIGINAL PAPER

Ant Lion Optimizer: A Comprehensive Survey of Its Variants

and Applications
Laith Abualigah1 · Mohammad Shehab2 · Mohammad Alshinwan1 · Seyedali Mirjalili3 · Mohamed Abd Elaziz4,5

Received: 15 July 2019 / Accepted: 9 March 2020 / Published online: 3 April 2020
© CIMNE, Barcelona, Spain 2020

Abstract
This paper introduces a comprehensive overview of the Ant Lion Optimizer (ALO). ALO is a novel metaheuristic swarm-
based approach introduced by Mirjalili in 2015 to emulate the hunting behavior of ant lions in nature life. The review is
highlighted the applications that are utilized ALO algorithm to solve various optimization problems. In ALO, the best solu-
tion is determined to enhance the performance of the functional and efficient during the optimization process by finding the
minimum or maximum values to solve a certain problem. Metaheuristic algorithms have become the focus of research due
to introduce of decision-making and asses the benefits in solving various optimization problems. Also, a review of ALO
variants is presented in this paper such as binary, modification, hybridization, enhanced, and others. The classifications of
the ALO’s applications include the benchmark functions, machine learning applications, networks applications, engineering
applications, software engineering, and Image processing. Finally, According to the reviewed papers published in the litera-
ture, the ALO algorithm is mostly utilized in solving various optimization problems. Presenting an overview and reviewing
the ALO applications are the main aims of this review paper.

Keywords Ant lion optimizer · Meta-heuristic optimization algorithms · Optimization problems · Nature-inspired
algorithms

1 Introduction searches [3], simulated annealing [4], hill climbing [5], and
𝛽 -hill climbing [6]. Nevertheless, the main benefit of these
Meta-heuristic optimization methods are mainly classified optimization methods is the accelerated search, while the
into two parts: local search-based optimization methods and main disadvantage is their aim focus on exploration (global)
population-based optimization methods [1]. Local search- search rather than exploitation (local) search, which, as
based optimization methods running with one solution dur- results, improves the probability of their getting local optima
ing its progress and it attempts to improve the single solu- trap. Counter to population-based optimization methods,
tion using the neighborhood mechanism [2], such as tabu which proceeding by using multi-solutions (population) at
a run, improving the candidate solutions to create one or
more better solutions at each generation. These optimization
* Laith Abualigah
aligah.2020@gmail.com methods are powerful in recognizing encouraging areas in
the wide search space but are still inefficient in exploiting
1
Faculty of Computer Sciences and Informatics, Amman the wide regions of the search space [7, 8].
Arab University, Amman, Jordan Optimization methods such as swarm intelligence and
2
Computer Science Department, Aqaba University evolutionary computation are classes of population-based
of Technology, Aqaba 77110, Jordan optimization methods [9]. These optimization methods are
3
Centre for Artificial Intelligence Research and Optimisation, based on the physical and natural biological evolution or
Torrens University Australia, Fortitude Valley, social communication operation of natural animals and oth-
Brisbane 4006, QLD, Australia
ers in life. Examples of swarm-based optimization methods
4
Department of Mathematics, Faculty of Science, Zagazig (algorithms) include Salp Swarm algorithm (SSA) [10],
University, Zagazig, Egypt
Particle Swarm Optimization (PSO) algorithm [11, 12],
5
School of Computer Science and Technology, Wuhan Grey Wolf Optimizer (GWO) algorithm [13], Gravitational
University of Technology, Wuhan, China

13
Vol.:(0123456789)
1398 L. Abualigah et al.

Search algorithm (GSA) [14], Moth-Flame Optimization discrete, binary, hybridization, and modified). Previously
(MFO) algorithm [15], Bat Algorithm (BA) [16], Artifi- published studies on the adjustment parameters of ALO are
cial Bee Colony (ABC) algorithm [17], Krill Herd (KH) additionally presented. We focus on the sources and prin-
algorithm [18], Krill Herd Algorithm (KHA) [19], Cuckoo ciples of ALO and its improvements and provide a com-
Search Algorithm (CSA) [20], and Ant Lion Optimization prehensive description of recent applications and connected
(ALO) [21, 22]. improvements accomplished over the last few years since
Swarm intelligence-based optimization methods are moti- 2015. At the conclusion of this review paper, readers will
vated by animal communities and social bug swarms. They be qualified to explain the design and the work principle
simulate the behavior of ant lions in their life cycle, which of ALO, as well as identify the developmental steps and
consists of two periods: larvae and adult [21]. ALO is a various fields that utilize ALO to find the optimal solutions.
recent meta-heuristic population-based optimization method Finally, the conclusion of this review paper is presented
proposed by Mirjalili in 2015 [15]. It is recognized in the with the advantages and weaknesses of ALO algorithm
lifestyle of ant lions which is the connexion of the volume and recommends possible future directions for work to the
of the snare and two points: the level of drought and appear- researchers who are interested. While the kinds of problems
ance of the moon. More details of the ALO algorithm are are generally examined the minimization, and some other
provided in [23]. kinds of problems are generally examined the maximization.
ALO algorithm combines the main two techniques (i.e., The authors will straight mention so when required. Figure 1
population-based search strategy and local-based search presents the number of published papers (i.e., conferences,
strategy) to produce an intelligent algorithm able to search journals, and chapter books), which are arranged based on
effectively by the two main search strategies (global explora- the publisher. It’s worth to mention that Fig. 1 shows the
tion and local exploitation). Related to other meta-heuristic number of published papers by several researchers from
algorithms, ALO algorithm is easy, adjustable and simply May-2015 to the first quarter of the year 2019 (May-2019)
implemented; as such, it can be employed to solve various on ALO algorithm as well as its variants according to the
variety of optimization problems [15]. Due to these charac- various report publishers such as IEEE, Elsevier, Springer,
ters, ALO are successfully utilized to several optimization and others. It is observed in Fig. 1 that in the last two years
problems, such as scheduling [24], feature selection [25, 26], (2017–2018) of ALO publications have exponentially
inverse problem and parameter estimation [27], networking grown. Figure 2 presents the distribution of these publica-
[28], classification [29], economic power dispatch planning tions based on the kinds of ALO applications. As well as,
[30], benchmark functions [31], software engineering [32], in Fig. 3, the distribution of these publications based on the
power energy [26], proportional integral derivative control- kinds of ALO variants.
ler [33], and image processing [34]. The rest of this review paper is organized as follows:
In this review paper, new ALO-based users are provided Sect. 2 introduces ALO algorithm by highlighting its struc-
with a comprehensive and exhaustive survey of the proce- ture and procedures. Section 3 presents an overview of the
dures of ALO algorithm and near readers with satisfactory ALO variants and developments. Application domains and
materials to know previously published variants (such as developments in particular areas are discussed in Sect. 4.

Fig. 1  Number of publications of Ant Lion Optimizer algorithm per year

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1399

The position of each ant in each dimension is updated

using a random walk. This random walk is as follows:
x(t) = [0, cumsum(2t(t1 ) − 1, cumsum(2t(t2 ) − 1, … ,
cumsum(2t(tT ) − 1] (2)
where T is the maximum number of iterations, ti shows the
tth iteration, cumsum is the cumulative summation, and r(t)
is a random function calculated as follows:
{
1 rand ≥ 0.5
r(t) =
0 rand < 0.5 (3)

where t indicates is the iteration index and rand is a ran-

domly generated number in [0, 1]
Fig. 2  Applications of the Ant Lion Optimizer algorithm To see how this random walk work, Fig. 4 is pro-
vided. This figure shows 50 random walks using the same

Section 5 discusses theoretical aspects, assessment, and

evaluation of Ant Lion Optimizer. Finally, Sect. 6 out-
lines some concluding remarks and outlines several future
research lines of interest.

2 Ant Lion Optimizer

Ant Lion Optimizer (ALO) mimics the way that ant lions dig
sand pits to hunt ants in nature.
In this algorithm, a population of solutions is first initial-
ized as ants on a search landscape. the position of each ant
is stored in a vector as follows:
[ ]
������⃗i = Ai,1 , Ai,1 , … , Ai,d
Ant (1)

where Anti shows ith ant, Ai,d shows the position of the ith
ant in the dth dimension.
Fig. 4  Fifty random walks using Eq. 2

Fig. 3  Variants of the Ant Lion

Optimizer algorithm

13
1400 L. Abualigah et al.

equation. It can be seen that the deviations can be quite �������������⃗1 ⎤

⎡ Antlion
abrupt starting from the first iteration. �������������⃗2 .⎥
⎢Antlion
As a population-based algorithm, ALO uses multiple ⎢ ⎥
MAntlion =⎢ . ⎥ (7)
ants and store them in a population matrix as follows: ⎢ . ⎥
⎢ �������������⃗ ⎥
�������⃗1 ⎤
⎡ Ant ⎣ Antlionn ⎦
�������⃗2 .⎥
⎢Ant
⎢ ⎥
Mant =⎢ . ⎥ (4) ⎡f (Antlion
�������������⃗1 )⎤
⎢ . ⎥ ⎢ ⎥
⎢ �������⃗ ⎥ �������������⃗2 )⎥
⎢f (Antlion
⎣ Antn ⎦ ⎢ . ⎥
Moal =⎢ ⎥ (8)
where n is the number of ants in the population. ⎢ . ⎥
In the ALO algorithms, seach agents (ants) are evalu- ⎢ . ⎥
⎢f (Antlion
�������������⃗n )⎥
ated using an objective function. The objective values are ⎣ ⎦
stored in the following vector:
where n is the number of ants in the population.
⎡f (Ant
�������⃗1 )⎤ In the ALO algorithm, ants move in a search landscape
⎢ ⎥ using the random walk presented above. All variables in
�������⃗2 )⎥
⎢f (Ant
⎢ . ⎥ ants are changed using diverse random walks. This mecha-
Moa =⎢ ⎥ (5) nism increases the diversity of the population in the ALO
⎢ . ⎥ algorithm. To keep the random walks inside the search
⎢ . ⎥
⎢f (Ant
�������⃗n )⎥ space and move them towards the most promising regions
⎣ ⎦ of the search space, they are normalized using the upper
and lower bounds of the search space and updated based
These equations allows calculating the positions and objec-
on the position of antlions. Better antlions have a higher
tive of ants as follows. Each ant represents a solution for
probability of impacting ant movements, which is done
a given optimization problem. However, they have to be
by a roulette wheel in ALO. There are two mechanisms
guided towards the promising regions of the search space.
that impact ants’ movements: one selected using a prob-
This is where the antlions come into play.
ability and the best antlion (elite). The range of all ran-
Each antlion is also represented with a position vector
dom walks are decreased proportional to the number of
and objective vector as follows:
iterations. Antlions are constantly updated in case of find-
[ ] ing an ant with better fitness. These mechanisms require
�������������⃗i = Ai,1 , Ai,1 , … , Ai,d
Antlion (6)
ALO to show superior performance on a large number of
where Antlioni shows ith antlion, Ai,d shows the position of optimization problems. To visually observe how the ALO
the ith ant in the dth dimension. algorithm solves optimization problems. Fig. 5 is given.

Fig. 5  Behaviour of the ALO algorithm on a unimodal and a multimodal test function

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1401

Fig. 4 shows that the test functions are unimodal (F1) 3.1 Basic Ant Lion Optimizer
and multimodal (F8). As can be seen in the search histories,
the exploration of the search space is reasonable in ALO. Basic ALO algorithm is proposed in [29] for the classifica-
The random walks in Fig. 4 are decreased proportional to tion problem of infinite impulse response (IIR) filters. As
the number of iterations, which causes a smooth transition mentioned above, ALO belongs to the population-based
between exploratory and explitative search patterns of solu- metaheuristic algorithms. Its working principle depends on
tions in ALO. The trajectory of solutions in Fig. 5 illustrate the cooperation between ants and antlions utilizing a tra-
high fluctuation that taper off proportional to the number ditional selection scheme called Roulette wheel selection
of iterations. Inspecting the average fitness of antlions and scheme for testing antlions to catch ants. During the optimi-
the covergence curves in Fig. 5, it is evident that the mech- zation processes, the performance of conventional antlion in
anisms of ALO improves the quality of solutions in each each round of process is matched with the best antlion which
iteratoin. This means that this algorithm is able to find a guaranteed an excellent solution. To confirm the effective-
reasonably good estimation for the global optimum of opti- ness of ALO in filter modeling problem, four different infi-
mization problems. nite impulse response datasets are examined for regular and
decreased ordered identifications. Modeling effectiveness
is evaluated by utilizing mean square error (MSE) among
3 Variants of Ant Lion Optimizer the identified and actual model achievements, mean square
deviation (MSD) between identified and actual different
This section describes the different variants of ALO algo- infinite impulse filter values and speed of convergence.
rithm in the following subsections (see Fig. 2). Table 1 illus- The proposed algorithm, ALO, bettered than the other two
trates a summary of the variants of the ALO algorithm in comparative algorithms (i.e., GSA and CSA). Furthermore,
solving optimization problems. the robustness of acquired solutions are evaluated, fifty

Table 1  Variants summary Category Problem Year References

of the Ant Lion Optimizer
algorithm Basic LabVIEW 2016 Gupta et al. [35]
ALO Linear discrete filters 2017 Nair et al. [29]
Healthcare Districting 2019 Farughi et al. [36]
Binary Engineering problems 2015 Mirjalili et al. [21]
ALO Benchmark Functions 2018 Gutierrez et al. [37]
Feature selection 2018 Mafarja et al. [26]
Modifications of ALO Finite metallic plates 2017 Chaleshtari et al. [38]
Unmanned aerial vehicle 2017 Yao et al. [39]
Parameter identification 2017 Wu et al. [27]
Radiation heat transfer 2017 Ksiazek et al. [40]
Weighted elitism 2017 Rajan et al. [41]
Optimal power flow 2017 Radha et al. [42]
Czochralski crystal growth 2018 Zhao et al. [43]
Wind Power Forecasting 2018 Jiang et al. [44]
Bearing Fault Diagnosis 2019 Yang et al. [45]
Hybridization of ALO Wind Power Forecaster 2017 Guo et al. [46]
Optimal cluster analysis 2018 Majhi et al. [47]
Micro-grid 2018 Fathy et al. [48]
Texture Classification 2019 Wang et al. [49]
Optimum design 2019 Azizi et al. [50]
Chaotic Optimizing SVM parameters 2016 Zhao et al. [51]
ALO Chaotic electroencephalogram Prediction 2018 Kose Utku [52]
Multi-objective of ALO Engineering problems 2017 Mirjalili et al. [53]
Micro-grid of renewable energies 2017 Hosseini et al. [54]
Distribution networks 2019 Abul’Wafa et al. [55]
Optimal Design 2019 Mousavifard et al. [56]
Economic Dispatch Problem 2019 Dhifaoui et al. [57]

13
1402 L. Abualigah et al.

independent identification cases are carry out and mean and a novel method for species migration is introduced. To
square error and mean square deviation rates are investigated assess the performance of algorithms, fifteen test functions
statistically using t-test on mean square error results. The are employed. Moreover, a comparison with other optimi-
proposed algorithm produced the smallest standard devia- zation algorithms, such as the PSO, GSA, GA, and BA is
tions showing best solutions. Additional, in t-test investiga- given. The proposed algorithms are further investigated for
tion, large positive t values newly showed the significant a real-world optimization problem.
superiority of the proposed algorithm over the comparatives Feature selection (FS) is represented as an optimization
for infinite impulse response in filter modeling. problem to determine the minimal features and the minimum
A new nature-inspired optimization algorithm, called data loss from an original dataset. FS problems belong to
Basic ALO toolkit, based on hunting way of antlions is the NP-hard problem. Thus, it is necessary to find a pow-
proposed in [35] for Lab-VIEW environment. An organized erful and rapid search algorithm to solve these problems.
procedure is provided for the improvement of the toolkit In [26], a new method has been introduced which merg-
containing a stochastic walk of ants, the construction of ing hill-climbing with binary ALO, namely, HBALO. The
nets, catching of prey and eventually reconstructing of authors starting their proposed method (i.e., HBALO) by
traps. Lab-VIEW is a rich development context employed creating the population (i.e., ants) randomly. Then, using
widely in judgment and handles applications. The only regu- two filters of FS to select the improved informative features
lar toolkit free in Lab-VIEW is Differential Evolution and in the collection. After that, achieving the optimal solution
consequently, an efficient optimizer to deal with this issue is by employing the HBALO. Evaluation the performance of
needed. The proposed method, ALO toolkit, is examined on the HBALO has been done by utilizing a group of eight-
various benchmark functions and it worked adequately well teen well-regard datasets from the repository of the UCI and
compared to the Differential Evolution toolkit. The obtained compared with the several related methods. The experiments
results proved that the proposed toolkit is suitable to produce results show that the superiority of the HBALO in exploring
superior results in regards to improved search, local optima the feature space for finding the optimal feature subset.
(exploitation) escape, and convergence rate.
The problem of population distracting in the good health 3.3 Modifications of Ant Lion Optimizer
system of South Khorasan region are examined on the one
hand of an optimization problem. Instantly, the distracting In this paper [39], a novel Dynamic Adaptive Ant Lion
problem is recognized as a vital matter, it is necessary to Optimizer (DAALO) for route planning of unmanned aerial
achieve dynamic solutions to perform in the system. There- vehicle. In this paper to make the ALO jump from local
fore, two optimization algorithms, ALO and GWO, are used optima more efficiently, the random walk of ants is replaced
in [36] to address a real-world problem (health care distract- by Levy flight. Furthermore, by including the development
ing). The objective function in this problem is to make as rate of the population as the feedback, the size of trap is set
max as possible (maximize) the population stability in each dynamically depending on the 1/5 Principle to enhance the
region. Problem limitations hold a unique task and the non- performance of ALO including meeting accuracy, meeting
existent distribution of abnormalities. Irregular allocation speed, and stability. The proposed algorithm shows supe-
indicates density, absence of contiguous, and lack of spaces rior performance to find the optimal route in two various
(holes) in the districts. Based on the achieved results, GWO environments such as city model and mountain model. In
has a better level of performance than the ALO. The out- addition, the result confirms the efficacy, robustness, and
come can be utilized as a beneficial tool for distracting in utility of DAALO.
other industries and domains of application. In the photovoltaic cell system, the tracking of the maxi-
mum power level and the forecast of power are very sensi-
3.2 Binary Ant Lion Optimizer tive. Therefore, determining the values and parameters must
be chosen accurately [38]. Wu et al. utilized the improved
Meta-heuristic algorithms are getting more important to ALO (IALO) to determine the first location of individuals
develop the solving processes and solve complicated prob- by chaotic order to improve the ergodicity and uniformity
lems that need a high computing capability. Biogeography- of population. The construct of the PSO is introduced in
based optimization (BBO) algorithm produced excellent determining the updated position of the individuals, where
results by utilizing a model of class migration within ecosys- the updated position based on the current best individuals
tems. Similarly, ALO can control the benchmark functions position and the global best individual. The IALO algorithm
very well when the decisions are continuous [21]. A major simulation shows better performance than other algorithms
modification for the two algorithms, in [37] to the BBO and like PSO, BA, and the standard ALO.
ALO, is how species migrate among ecosystems by combin- Another topic is the engineering processes are of supreme
ing two features, migration rules are randomly determined, significance for the economy and industry. For this sake,

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1403

we need effective simulation and positioning systems that the generalization and convergence rate ability of BP neural
will efficiently maintain human operators. A radiation heat network. The model is examined by collected data of an Irish
transfer by using electric furnace is one of these problems. In wind farm in November 2017. The tested results demonstrate
this study [40], proposed simulation and positioning system that the proposed algorithm can succeed the drawback that
of its operation parameters by the using modified ALO algo- the original algorithm is easy to fall into the convergence
rithm. The result demonstrates that the proposed approach speed is slow and the local optimum. Besides, the IALO-
can efficiently position the system. BP algorithm is better to BP neural network, Generalized
The power systems tackling a problem named the Opti- Regression Neural Network (GRNN), and Support Vector
mal Reactive Power Dispatch (ORPD). Thus, Rajan et al. Regression (SVR) in the prediction efficiency and stability.
proposed to applied the ALO algorithm [41]. The ORPD This study [42], intends to reduce the costs, observe to
is a reactive power (VAr) organizing problem, non-convex, prevision norms and preserve supply and voltage points sub-
non-linear, and a challenging optimization problem. In addi- ject to various operational issues, containing fuel limitation
tion, the authors worked to develop the searchability of ALO matters and non-linear properties of generators, which are
by utilized weighted elitism method. That’s lead to improve beneficial for the existing power system. A developed con-
the global search of the algorithm. The suggested Modified trol framework is required while regarding multiple different
ALO (MALO) effectively supports each of exploitation and operational goals that require utilizing the ALO as a major
exploration, which improves the hunting ability of ALO. The optimization technique. Fuzzy algorithm is included in ALO
MALO and the original ALO are applied to find the tap loca- to obtain the best compromise solution (BCS) amongst a set
tions, VAr product of shunt capacitors, and optimal settings of non-dominated solutions. Simulation results show that the
of generator voltages to optimize purposes. For instance, ability of the proposed method to find the multi-objective
the active power loss, voltage stability index, and the total dynamic optimal power flow (DOPF).
voltage variation. The MATLAB is used to simulate the pro-
posed method. Then using IEEE-30 and 57 bus standard 3.4 Hybridizations of Ant Lion Optimizer
systems to evaluate the performance. The experiment results
illustrated comparisons between ALO and other algorithms; Wind power has the properties of evaporation and intermit-
(i) a set of benchmark functions has been used to improve tency. Thus, its entrance into the field of the electric power
the efficiency of the MALO. (ii) The t-Test and paramet- systems increases the defies in terms of the safe and steady
ric analysis approve the consistency and robustness of the operation, which leads to making accurate wind power fore-
MALO compared with ALO and other similar systems. casting important, extremely. Consequently, Guo et al. pro-
In this paper [43], propose an enhanced ALO algorithm posed a hybrid wind power forecasting system using ALO,
with Levy mutation and adaptive elite competition tech- namely, BND-ALO-RVM forecaster [46]. The proposed
nique. The random number following Levy dissemination method mixes several algorithms; The relevance vector
is applied to mutate poor individuals in the population, machine (RVM), Beveridge-Nelson decomposition method
which can increase the variety of population to develop (BND) and ant lion optimizer (ALO). Regarding the non-
global search capability of the algorithm. Furthermore, stability and nonlinear components of wind power infor-
adaptive elite competition technique that several elites drive mation, the wind power time series disintegrated into sto-
the population to explore simultaneously can facilitate the chastic, deterministic, and cyclical elements utilizing BND
convergence speed. To decrease the number of calculation, method. After that, forecasted these three parts individually
the number of elites will reduce with the rise of repetitions. by employing RVM method. In the meanwhile, ALO using
The performance of the proposed algorithm is compared to find the kernel width parameter of RVM for enhancing
with the ALO algorithm and Particle Swarm Optimization the forecasting execution. The outcome was calculated by
(PSO). The results demonstrate that optimization accuracy multiplying the forecasting results of the above three ele-
and convergence speed of the improved ALO algorithm is ments. The BND-ALO-RVM was tested with real wind
better. The goal of the improved ALO algorithm is applied power data from the Xinjiang Uygur region in China. The
to the identification of thermic field temperature method in experiment results showed that the proposed BND-ALO-
Czochralski silicon single crystal increase and tested results RVM forecaster has the best forecasting performance of all
show the effectiveness of the proposed approach. the examined methods, which confirms its efficacy.
In this research [44], a modified ALO algorithm to opti- In this paper [47], a hybrid clustering method based on
mize the back propagation (BP) neural network (IALO- the K-means algorithm and ALO algorithm proposed for
BP) is proposed to enhance the prediction efficiency of optimal cluster analysis. To evaluate the performance of
short-term wind power. The proposed model applies the the proposed model the result compared with K-Means,
improved ALO algorithm (IALO) to optimize the thresh- KMeans-PSO, DBSCAN, Revised DBSCAN and KMeans-
olds and weights of BP neural network, so as to enhance FA clustering algorithms based on various performance

13
1404 L. Abualigah et al.

metrics. The experimental results are examined on eight theory is proposed in [51], which attracts experience from
datasets, for which the statistical analysis is accomplished. ABC and carries in chaos search based on the information
The result demonstrates that the proposed hybrid of K-means of ALO. The ALO with chaotic investigation theory algo-
algorithm and ALO algorithm show a superior performance rithm firstly outlines poor-fitness solutions. Meantime, the
than the other three models in terms of the sum of intraclus- position of these solutions is considered as the initial value
ter distances and F-measure. of chaotic maps. Then it can obtain excellent results by a
Fathy and Abdelaziz proposed a new algorithm to man- specific number of chaos iterations and reappoint the posi-
age the procedure of the microgrid (MG) with established tion to ALO, which is useful to enhance the excellence of the
renewable energy sources (RESs), by using ALO and KH ALO population and its performance. Finally, the ALO with
optimization algorithms [48]. The KH is utilized for fixing chaotic investigation theory algorithm is employed to find
the single-objective function describes whole procedure cost the optimal parameters of the SVM. The well-known bench-
or whole pollutant emission included from the established mark datasets are collected from the University of California
producing units. On the other hand, ALO is utilized to fix Irvine (UCI) and employed for assessing the proposed ALO
the multi-objective function of whole procedure cost and with chaotic investigation theory algorithm. The obtained
emission together. The problem is formed as a non-linear results indicated that the ALO with chaotic investigation
constrained objective function with balance and variation theory algorithm for optimizing the parameters of SVM got
chains. In this work; the current machines in MGs are wind better results in terms of its performance and stability.
turbine (WT), fuel cell (FC), microturbine (MT), photovol-
taic panel (PV), grid and battery. The authors used two sce- 3.6 Multi‑objective Ant Lion Optimizer
narios for the evaluation: (i) Optimizing MG with setting up
all RESs within particularized destinations. (ii) Operating A multi-objective ALO algorithm is proposed in [53] as
both WT and PV at their measured powers. The collected a new version (MOALO). A repository is leading used to
results are compared with various algorithms like Fuzzy save non-dominated Pareto current optimal solutions. Solu-
self-adaptive PSO (FSAPSO), genetic algorithm (GA), grey- tions are later taken from this repository utilizing a selection
wolf optimizer (GWO), particle swarm optimization (PSO) mechanism (e.e., roulette wheel) according to the coverage
and whale optimization algorithm (WOA). rate of solutions to lead ants towards richness regions of
multi-objective exploration spaces. To confirm the perfor-
mance of the proposed algorithm, a collection of common
3.5 Chaotic Ant Lion Optimizer unconstrained and constrained benchmark test functions is
used. Furthermore, the algorithm is utilized to a kind of
The forecast of future events based on possible time series multi-objective engineering problems: brushless dc wheel
analyses is an important research domain especially for motor design, cantilever beam design, disk brake design,
health-care problems, such as forecast and assessments of 4-bar truss design, security isolating transformer device,
attack purposes. A test of brain dynamics is routinely inves- the rate decreased design and welded beam pattern. The
tigated to get information about course, as well as future, obtained results are tested by comparing the multi-objective
mental events, and to discover and diagnose infections. ALO algorithm against NSGA-II and multi-objective PSO
Because of their chaotic energy, electroencephalogram time algorithm. The results of the proposed multi-objective ALO
series need specific techniques for a powerful forecast. A algorithm on the test functions proved that this algorithm
hybrid method is developed in [52] through AI techniques to exhibited a high convergence rate. The obtained results on
address with electroencephalogram time series. Both ALO the engineering problems proved the applicability of the
and artificial neural networks, which is a new intelligent proposed multi-objective ALO algorithm is able to solve
technique, were used to understand the related system and challenging real-world optimization problems as well.
implement forecast applications over electroencephalogram A sustainable development simulation method is pro-
time series. Based on the obtained results, the system suc- posed in [54] for controlling energy supplies in regards to the
cessfully forecasted the future states of purpose time series virtual power players working in a smart-grid. The proposed
and it exceeded several other hybrid systems and various power resource control schedule in a micro-grid, containing
time series forecast approaches published in the literature. fuel cells, solar panels, micro-turbines, wind turbines, and
ALO is a novel technique having artificial parts for the batteries, effectively faces the demands of the grid. Out of
improvement and development process. While the popula- utilizing the above-mentioned devices, virtual power players
tion of basic ALO has somewhat poor-fitness solutions, the can further purchase more power from upper use to return
reaction of ants choosing those ant lions for the chance step to the load. In extension, the proposed method represented
to increase the likelihood of its stuck in local optima and to fit a micro-grid utilizing a multi-objective frame, which
changing its performance. ALO with chaotic investigation reduces the total operating expense and radiation affected by

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1405

the generating parts concurrently. To accomplish this aim, Table 2  Applications summary of the Ant Lion Optimizer algorithm
the multi-objective ALO algorithm are employed to address Category Date References
the multi-objective problem and to generate Pareto optimal
solutions. The fuzzy method is applied for the judgment Benchmark functions 2017 Dinkar et al. [31]
making the process. Eventually, to prove the performance 2018 Wang et al. [58]
of the proposed multi-objective ALO algorithm, the results Machine learning 2018 Hu et al. [59]
are matched with multi-objective PSO algorithm and Non- 2019 Wang et al. [34]
dominated Sorting GA, which proved that the performance 2020 Heidari et al. [60]
of the multi-objective ALO algorithm in the presence of Engineering applications
fuzzy way achieves the better solutions in terms of emission Scheduling 2016 Tung et al. [24]
of pollutant and operation cost. 2016 Tung et al. [24]
Optimal concurrent placement of categorized generators 2016 Umamaheswari et al. [61]
and contemporary condensers (DGs-SCs) is arranged in [55] 2017 Umamaheswari et al. [62]
utilizing ALO algorithm based multi-objective optimization, 2017 Umamaheswari et al. [63]
recognizing current and reactive energy waste and voltage 2018 Silberschatz et al. [64]
resistance as purposes. The multi-objective formulation is 2018 Elango et al. [65]
transformed into a sole objective from the weighted total 2019 Utama et al. [66]
of the specific objectives. A synchronous condenser is rec- 2019 Dinkar et al. [67]
ognized due to its capacity to receive or provide reactive Control of power systems 2016 Trivedi et al. [68]
energy. A capacitor bank does not have this ability. The 2016 Engel et al. [69]
performance of simultaneous placement of PVDGs with 2016 Trivedi et al. [70]
SCs, Wind-based DGs with SCs, and micro-gas turbine 2016 Parmar et al. [71]
DGs (uTDGs) with synchronous condensers are examined 2017 Yasin et al. [72]
on IEEE 118-bus long-scale radial delivery system (RDS). 2017 Mansour et al. [73]
Furthermore, the effectiveness of simultaneous station of 2018 Hatata et al. [74]
the different renewable DGs: PV, Wind based, uT, with SCs 2018 Spoljari’c et al. [75]
are examined on the same RDS. A limited yearly investment 2018 Filho et al. [76]
cost per unit power damage loss of optimally designated 2018 Hatata et al. [77]
several DG and SC is introduced. 2018 Ali et al. [78]
Economic load dispatch prob- 2016 Mei et al. [79]
lem 2016 Balachandar et al. [80]
4 Applications of Ant Lion Optimizer 2016 Mouassa et al. [30]
2016 Kamboj et al. [81]
ALO algorithm is employed effectively to address applica- 2016 Mei et al. [82]
tions of many different disciplines such as benchmark func- 2016 J’unior et al. [83]
tion, machine learning, engineering design problems, and 2018 J’unior et al. [84]
others. Expressing proper problem and determining suitable 2019 Alazemi et al. [85]
variables and objective functions are the important matters 2019 Hatata et al. [86]
that can assist effective problem solving when solving opti- PID controller 2016 Saikia et al. [87]
mization problems. The conventional utility of operators, 2016 Prasad et al. [88]
parameters adjustment, and population model are the major 2017 Pradhan et al. [89]
issues as respects optimization algorithms. In this section, a 2018 Mokeddem et al. [90]
study of the literature is conducted by classifying the appli- 2018 Spoljari’c et al. [91]
cation (scientific applicability) domains in the following 2018 Pradhan et al. [92]
subsections (see Fig. 3). Table 2 shows a summary of the 2018 Kanimozhi et al. [93]
applications of the ALO algorithm in solving optimization 2019 Eltag et al. [33]
problems. 2019 Srinivasan et al. [94]
Optimal design problem 2016 et al. [95]
4.1 Benchmark Functions 2019 Chaleshtari et al. [96]
Smart Grid 2017 Sam’on et al. [97]
Effectiveness of any optimization algorithm, nature-inspired, 2019 Roy et al. [98]
is controlled by appropriate operators employed for perform- Image processing 2019 Leke et al. [99]
ing the search strategies (i.e., exploration/diversification and

13
1406 L. Abualigah et al.

Table 2  (continued) 4.2 Machine Learning Applications

Category Date References
ALO show superior performance in many applications, how-
Software engineering 2018 Rayyam et al. [32] ever, its population variety and concentration accuracy can
2019 Sharma et al. [100] be limited in some applications. [59] proposed a method
Networks applications 2016 Zainal et al. [28] using the ALO algorithm and the common genetic operator
2018 Kaur et al. [101] (i.e., mutation operator) named ALO-DM. In this method,
2016 Maher et al. [102] a tight approach and differential mutation operator improve
2016 George et al. [103] the variety of the population. Moreover, merging it with
data-mining approaches can be beneficial and effective in
big data analytics difficulties. The obtained results demon-
exploitatio/intensification search strategies). The absence strated that the proposed approach is effective in finding a
of these strategies tends the underlying algorithm towards feasible, efficient, and accurate solution.
early convergence, trap of local optima and failure to find In the hyperspectral image (HSI) classification the fea-
and obtain the global optimal solution. A new novel opti- ture selection is the most important issues to improve the
mization algorithm, opposition based ALO algorithm, is interconnection between the neighboring chains. The goal is
proposed in [31] to get better performance in the process of finding fewer chains with the greatest exactness as possible.
the ALO algorithm. For gaining acceleration in the search However, it is a complicated NP-hard problem and cannot
process, exploration search strategy is to be improved. A be addressed by evolutionary algorithms. Therefore, in [34]
couple of approaches are employed for reaching this aim: proposes to use a combination of AOL algorithm, Wavelet
Firstly, Laplace arrangement is utilized in a stochastic walk SVM to improve the fastness of the classification outcome,
(random) of ALO algorithm rather than stable distribution and Lévy flight assists optimization algorithms standing out
which guarantees the diversification of wider search region of the best local optimum. A new feature selection algorithm
than the initial stochastic walk of ALO algorithm. Secondly, is proposed, a modified version of the ALO algorithm and
the Opposition-based Learning design which guarantees the W-SVM (MALO), to decrease the magnitude of HSIs. The
diversification search of candidate solutions in the search proposed algorithm is compared with the well-known algo-
region simultaneously to determine the better candidate rithms on some hyperspectral image datasets. Furthermore,a
solutions when the development process is in running. A new evaluation criterion is formed to evaluate the classifica-
full set of twenty-seven benchmark functions including a tion accuracy, and the performance of feature selection and a
wide variety of various characteristics and various dimen- chosen number of chains are balanced as far as conceivable.
sions are used for verifying of results. Moreover, the impact The result shows that the suggested approach has higher
of Laplace distribution and opposition-based new solutions performance matched with other approaches in regards of
production while the development process is investigated by finds the optimal solution with a fair concentration adjust-
the performance of trajectories, convergence speed, objec- ment, and its classification efficiency is supplied with fewer
tive function values utilizing boxplot and the improvement chains, it is strong, and could be employed for effective new
values of average fitness functions for each test group. The work in the problem of the features selection.
proposed opposition based ALO algorithm is further tested An efficient training technique is introduced in [60]
using the set of un-constrained engineering problems of based on the ALO algorithm to solve the Multi-Layer neu-
Optimal Capacity of Gas Production Facilities and Gear ral networks problem. ALO is a well-known optimization
Train Design, revealing heterogeneity in solving several algorithm motivated by the sharp chasing of antlions in life.
real-world problems. First, the technical experiences of ALO are fully illustrated.
Proposing at the weakness that the ALO algorithm has Second, brief literature and history is presented according to
unbalanced exploration search and improvement ability, an modern established researches. Moreover, a useful encoding
improved ALO algorithm with adaptive boundary and opti- plan is given and the objective formula is represented, math-
mal leadership is introduced [58]. First, the ALO algorithm ematically. The introduced training approach using the basic
tuning the range of the border to attain equilibrium in the ALO algorithm is tested on sixteen common benchmarks.
diversification search and improvement abilities. Second, The performance of the proposed training model using the
to enhance the convergence acceleration and global search basic ALO is compared with other algorithms including
capability, the optimal-guided comparison is adapted. The DE, GA, PSO, and population-based incremental learning
obtained results of six test benchmark functions revealed in regards to the accuracy values (i.e., best, worst, average,
that the improved ALO algorithm enhanced the efficiency and median). Besides, convergence abilities are analyzed for
and convergence rate of the optimal solution analyzed with all comparative algorithms. The experiments showed that the
other well-known optimization algorithms. introduced training model obtained better results compared

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1407

to GA, population-based learning, DE, and PSO in analyz- tasks [24]. Three CPU scheduling policies have been used to
ing the majority of used datasets and produced enhanced compare with the performance of ALO: round robin (RR),
accuracy values and acceleration rates. shortest computation time first (SCTF), and first come first
serve (FCFS).
In [65], the authors focused on generator maintenance
4.3 Engineering Applications scheduling (GMS) because of its importance in improving
the generator’s performance and increasing its lifetime. For
4.3.1 Scheduling that, they used ALO to address the preventive GMS problem
in energy systems. The objectives of the work summarized
Optimization problems that include determining resources in minimizing risk, total operational, and maintenance costs.
to execute the tasks at specific times are called scheduling Also, maximizing system reliability. One-way ANOVA and
problems. Such problems aim to determine a schedule for Wilcoxon signed rank tests have been utilized to validate the
minimizing the total of the amount of time or the cost (or effectiveness of the introduced work.
both) that is required to complete all the tasks.
Integrated Maintenance Scheduling (IMS) is very nec- 4.3.2 Control of Power Systems
essary to save the systems’ continuation and performance.
However, there are some challenges such as the outages Ekaterina et al. adopted ALO algorithm to optimize the
leads to interruption of communication between the trans- power location of a partly shaded photovoltaic module [69].
mission lines and the generating units. Thus, Umamaheswari Trivedi et al. used ALO to solve the power system engineer-
et al. utilized ALO for addressing the IMS problem [63]. ing problems [68]. For instance, determine the minimum
The servicing schedules of power elements were dealt as active energy loss, fuel cost, and reactive energy loss. In
elements and the fitness value is effective or productive addition, ALO helps to coordinate the overvoltage trans-
in relation to IMS cost. The authors used the IU-19 unit, mission and improve the efficiency of the isolated safety
IEEE-30 bus, and IEEE-RTS to assess the effectiveness of transformer. The authors utilized electrical datasets to assess
ALO. The achieved results showed that the ALO proved its the effectiveness of ALO. The obtained results showed that
efficiency. The same authors used ALO to the preventive ALO outperformed PSO and Firefly Algorithm (FA) by
servicing scheduling of creating units based on deterministic achieving less computational time and high effectiveness
precision [61, 62]. It’s worth to mention that deterministic from the resource.
reliability represents the actual situation of preventive main- The ALO algorithm and its variants have been used to
tenance scheduling. The main purpose of employing ALO enhance the quality of different egergy systems. For exam-
is determining the minimum value of the deterministic reli- ple, Hatata and Hafez [74] proposed the ALO as an alter-
ability model which called Sum of Squares of Reserve (SSR) native method for solving Economic load dispatch (ELD)
through the scheduling times. which incorporates the Demand response as virtual power
Tung and Chakravorty used ALO to find the optimal tabu- plant. The main aim of this algorithm is to find the minimum
lation of thermal parts for small scale electrical economic value of the overall cost, system losses, and emission levels.
power dispatch planning (EEPDP) [24]. The work included The ALO is compared with other two meta-heuristics meth-
three main points (i.e., to find the optimal); (i) Determine the ods, namely PSO and artificial immune system (AIS), using
variables such as power generation (PG) which is the start- six-bus, thirteen-bus, and one hundred eighteen- bus IEEE
ing point for the initial solution in ALO. Also, the objective systems. The experimental results proved that the introduced
function which represents the fuel cost. (ii) Constraints, for algorithm got the high performance among the comparative
instance, the constraints variables of PG should be incor- algorithms (PSO and AIS), in addition established that the
porated in ALO. (iii) Stopping criteria, which refer to the incorporation DR in ELD decreases the costs/losses/emis-
maximum iteration limit to find the optimum solution. Three sions without changing the energy system protection and
and six unit system have been used in the tests to assess the authenticity.
effectiveness of ALO. The outcomes showed the ALO out- In [70], the ALO is applied to improve the quality of
performed seven algorithms applied in the literature. another power system called optimal power flow (OPF). In
CPU scheduling could affect performance metrics which this study, the ALO is compared with PSO and Firefly Algo-
defined as average turnaround time, response time, or CPU rithm (FA) to solve OPF using the IEEE-30 bus as tested
utilization [64]. The main objective of CPU scheduling is system. The three algorithms are aimed to solve three prob-
improving the system’s performance by minimizing the lems called Fuel Cost Reduction, Voltage Deviation Mini-
process waiting time and increasing throughput. Dinkar mization, and Voltage Stability amelioration. Based on the
and Deep employed ALO for enhancing CPU scheduling comparison results the ALO obtained better results in com-
task by determining the minimum average waiting time of parisons with the other two MH methods.

13
1408 L. Abualigah et al.

As well as, the ALO is used to enhance the fineness of which include the PSO, salp swarm algorithm (SSA) and
the solution of the OPF but with different objectives is intro- velocity relaxed PSO. ALL of these methods use the mean
duced in [71]. These objectives are Minimize the Reactive cost of time region transitional process fineness indicators
Power Losses with minimization the Fuel Cost Reduction, and integral of time-weighted absolute error (ITAE) as
and Active Power Losses. The proposed OPF approach fitness functions.
based on ALO is compared with, also, FA and PSO. In addi- [76] presents the ALO to determine the optimal solu-
tion, the authors provides a set of different mono-objective tion for PSS and the energy oscillation damping (POD)
problems which simulated by non-quadratic total gas emis- handler of the resilient AC transmission system (FACTS)
sion, fuel cost (with smooth and non-smooth functions), and devices with existing the remote signals (RSs) in multi-
total active losses. In this study, the IEEE-30 bus test system machine energy systems. The Pade approximation is
is utilized and the results of ALO is compared with differ- modeled the RSs which applied for the distributed energy
ent MH methods for example, when applied to solve total resources damping of inter-area vibration modes. The most
gas emission the hybrid Firefly Algorithm and others algo- two FACTS namely thyristor-controlled series capacitor
rithms. Among all the tested problem the ALO established and the static var compensator are used. For tunning the
its performance overall the others methods. PSS and FACTS-PODS, the New England-New York six-
Zuhaila et al. [72], used the ALO to find the optimal solu- teen-generator sixty-eight-bus system, and the Brazilian
tion to the Undervoltage Load Shedding (UVLS) problem. equivalent method formed with twenty-four simultaneous
The voltage stability (VS) index is used to determine the machines and one hundred seven-buses are used as appli-
load’s location to be shed since the VS index is a good meas- cations. The results of the ALO is compared with PSO and
ure to control the critical bus. The main aims of using the sequential quadratic programming and it provides better
ALO is to find the best result of the load to be dropped for results over both of them.
minimizing the energy loss and improved the voltage profile. Three strong optimization techniques (ALO, artificial
In this proposed approach uses the IEEE 30-bus as tested immune system, and PSO) are compared in [77] to opti-
system. The comparison results illustrate that the ALO has mize multi power sources in a distributed system. The
better performance according to power loss minimization, ALO is also applied to another application called distrib-
voltage profile improvement and stability improvement. uted generations (DG) in distribution system. Where the
In [73], ALO is used to determine the suitable spread ALO is this application aims to find solution which mini-
power resources allocation in different radial distribution mizes the whole running time of the directional over cur-
network which aims to maximize the percentage of power rent (DOC) relays. The authors used the practical 11-bus
losses reduction. To assess the production of the introduced distribution system and IEEE-30 bus (distribution section)
ALO, it is matched with a set of different methods includ- from Egyptian West Delta Distribution Utility as tested
ing chaotic symbiotic organisms search, quasi-oppositional systems. The experimental results proved the high per-
instruction learning-based optimization technique, cuckoo formance of the ALO with reducing the execution time
search, simulated annealing, grid search, and oppositional over the PSO and AIS methods. In addition, to determine
gravitational search. This comparison is performed using the optimal location and DG units in Radial Distribution
different number of installed DERs with two different sys- Networks (RDNs), the authors of [78] used ALO. In this
tems namely IEEE 34-bus and 118-bus radial distribution method, the 33, and 69 standard bus systems are used with
networks. According the comparison results the ALO veri- the Photovoltaic PV blocks are utilized as DG systems
fied its superiority overall the other methods. which aims to reduce the total active losses and increase
Moreover, the ALO has been applied to harmonious the voltage status of all operation buses. This problem
adjustment of power system stabilizers (PSS). For exam- is recognized as a multi-objective optimization dilemma.
ple, the authors in [75] proposed a method using on ALO The algorithm established its performance among other
for adjusting excitation controls in multi-machine energy optimization techniques.
system. This tuning contains the parameters change in
the PSS and automatic voltage regulator (AVR). The 4.3.3 Economic Load Dispatch Problem
two interconnected transmission lines with a two area
with four machine model (TAFM) are used to assess the ALO helps researchers analyze and predict the market to
effectiveness of ALO to determine the optimal solution reach optimal gains. ALO has been used in many economic
for solving the problems of small signal stability. Those aspects, such as predicting stock price, electricity price fore-
problems introduce automatic re-closure of the linkage casting, fossil-fuelled power plant, optimizing fuel consump-
position between two regions, a short line on the linkage, tion and load/dispatch problems. Mei et al. utilizing ALO to
short term load interruption in the first region, and a load address optimal reactive power dispatch (ORPD) problem
change in first area. The ALO superiors the other methods through determining the optimal set of control variables

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1409

incorporate the generator voltages, reactive compensation ALO to regulate the performance of PID Controller for
devices, and transformer tap ratios. Thus, minimize the an Automobile Cruise Control System (ACCS) [89]. The
transmission losses. IEEE 30-bus system has been utilized to aim of using ALO is designing the ACCS problems such
assess the effectiveness of ALO, where the obtained results as rise time, settling time, peak time, and maximum over-
showed that it is an efficient method in solving ORPD prob- shoot. Thus it helps to determine the optimal PID controller
lems [79, 80, 82]. Also, Mouassa et al. used ALO to solve parameters as per the specification of the scheme, where the
the same problem (i.e., ORPD) [30]. The authors achieved fitness function is determining the minimum time response
the best control variables of ORPD. For instance, the quan- which obtained from the dependency of an error on time.
tity of switchable capacitors stores, the location of tap Raju et al. applied ALO algorithm to adress the differ-
changers of transformers, and generators terminal voltage. ent gains of PID + Second Order Derivative (DD) control-
In the evaluation, they utilized IEEE thirteen-bus, IEEE one ler problem [87]. The authors summarized the goals of the
hundred eighteen-bus, and IEEE three hundred-bus power work in three points; (1) Utilized the ALO find the opti-
systems. mal controller gains different controllers (i.e., Integral (I),
This technique, economic load dispatch problem (ELDP), Proportional-Integral (PI), PID, and PID+DD). (2) Analyze
allows the reactive and real power of the generators to shift the best controller in inertia constant, system loading, posi-
in a wide area which leads to achieving a load demand in tion and size of disturbance. (3) Evaluate the performance
lowest cost. Kamboj et al. used the ALO to solve the non- of the best controller using the Random Loading Pattern
convex and changing (dynamic) ELDP of the electric energy (RLP). The simulation results proved that the production of
system [81]. The authors employed ALO to help the ELDP the introduced method, ALO with the PID + DD controller,
in defining the optimal operating cost (minimum fuel cost), obtained better results compared to PID controller.
where it determines the boundaries values, with knowing ALO based Fractional Order PID (FOPID) handler for
that there is a penalty for any violated. Finally, four IEEE speed controller and reduce the torque ripple of Switched
benchmarks have been tested to evaluate the performance Reluctance Motor (SRM) drive system is proposed in [88].
of ALO. The authors take the advantages of ALO to find the opti-
Economic emission load dispatch (EELD) is used in order mal control signal by tuning the FOPID gain parameters for
to reduce the total cost of carbon emissions and fuel con- supporting the SRM. Consequently, torque, speed, and the
suming. Therefore, both of [83] and [84] focused on solving performance of SRM will be determined. The evaluation of
the EELD by employing ALO. There is a multi-objective the performance of ALO achieved better results matched to
(i.e., emission function and cost function) has been used the PSO algorithm and genetic algorithm (GA).
with their respective restrictions. IEEE bus test system has Mokeddem and Draidi employed the ALO to find optimal
been applied to evaluate the efficiency of ALO, where the PID sliding surface parameters with the sliding mode con-
results of the experiments proved its efficiency in generation trol (SMC) system [90]. The proposed approach provides a
and reduction of the total emission. strong sliding mode controller and authorizing the system to
reach minimum chattering and maximum switching. Lyapu-
4.3.4 Optimal Design Problem nov theorem has been used to evaluate the effectiveness of
the tuning of PID sliding surface using ALO.
The engineering design procedure is a set of actions that In [91, 92], the ALO has been applied for determining
designers tend to achieve a solution for solving the given the optimal parameters of the Proportional Integral Deri-
design problem, and usually, the solution includes produc- vate (PID) controller for an Automatic Voltage Regulator
ing a result (e.g. design product such as a computer code or (AVR) system. The ALO is compared with different set of
a machine) that satisfies specified standards and/or achieves algorithm which it has been found that the accuracy of the
a particular task. Talatahari utilized ALO to determine the ALO is better than most of the other methods. Meanwhile,
optimal (i.e., optimum design of truss structures) design a modified version of the ALO is utilized to determine the
of the skeletal structures [95]. The author tested the per- coefficients values of the single diode (SD) approach of the
formance of the ALO by designing three truss and frame Photovoltaic (PV) solar cells [93]. This modified version
structures. After that, compared it with other meta-heuristic using the LambertW function under the effect of irradiance,
algorithms in structural optimization. heat, cold, and warmth (temperature) during the inves-
tigation of the I-V data with Root Mean Square Error as
4.3.5 Proportional Integral Derivative (PID) Controller objective function. In this study, the introduced algorithm
is compared with the most popular methods utilized to find
The Proportional Integral Derivative (PID) handler is the the coefficients of SD approach for PV and its provide com-
common popular power way, which based on a control loop petitive results with them.
feedback mechanism [33, 94]. Pradhan et al. introduced

13
1410 L. Abualigah et al.

4.3.6 Smart Grid industry. The high functionalities of object-orientation not

only enhance the improved software but also include a wide
The ALO is used as an alternative technique for energy variety of problems in a testing (measurement). For exam-
management system (EMS) as presented in [98]. Where the ple, not just the object but also the difficulty of object influ-
Recurrent Neural Network (RNN) with aid of ALO has been ences testing. Several other serious problems of testing come
applied to determine the power arrangement in Microgrid along with object-orientation. Consequently, this is always
(MG) framework. The proposed approach aims to minimize the main concern of software testers. Therefore, it is very
the electricity production cost by a permanently day-ahead critical to optimize this whole rule of object-oriented regard-
and real-time tabulation with the economic time-dependent ing software testing. Various nature-inspired optimization
constraints. Also, the demand response (DR) is assessed by algorithms are employed in the history record to achieve this
using the RNN and other measures to evaluate customer goal. A novel optimization algorithm, ALO, is introduced in
response, for example, DR magnitude, consumer’s report [100] to find the optimal test tracks for object-oriented soft-
based on the given preference, minimum cost of energy ware. Hither, in the proposed method, model-based testing
(COE), and duration. In addition, the ALO algorithm is is conducted rather than code-based testing as this assists
adapted to find the solution of the economic dispatch with in decreasing the cost and testing effort by recognizing the
finding the generation, storage, and sensitive load offers. The bugs and error in the design stage itself. The aim of the
introduced approach is matched with GA, ABC, and BFA to introduced model-based software testing is to include all
assess the performance of that approach. the transitions being in the system (i.e., full track coverage
In [97], the ALO is applied for addressing unit commit- must be beyond). A random factor is presented in the fit-
ment Problem (UCP) in Smart Grid System with uncertain- ness function to gain the random movement of ant lions in
ties. The ten-unit ways with the limitations, such as genera- the search area in the exploration of food. ALO is used on
tion limit, spinning resources, power balance, the minimum three applications of object-oriented software and the result
level and downtime restrictions are used to establish the per- ensures full track coverage which is a master purpose of
formance of the AL. The results of ALO are matched with software-testing.
the Dynamic Programming (DP) and they provide evidence To enhance the sensories handler of judgment motor uti-
about the using of ALO and the integration of solar power lizing Kalman filtering (UKF) problem, a new meta-heu-
resources in UCP scheduling can enhance the total running ristic optimizer, called ALO algorithm, is proposed in [32]
costs significantly. for online rotor acceleration and flow estimation. The main
problem with UKF detective is its sensitivity to the first val-
4.4 Image Processing ues of Q and R. To gain the optimal solution of these models,
a novel method called ALO-UKF is proposed. It is accord-
ALO algorithm is an evolutionary optimization algorithm ing to the compound of the standard UKF detective and the
able to determine near-optimal solutions to various complex ALO algorithm. Integrated ALO-UKF has yielded excellent
optimization problems. A different unique framework for the outcomes compared with the common Kalman filter and the
ALO algorithm is presented in [99] to avoid data imputation original UKF detective in regards to efficiency measure. A
in the problems of high-dimensional datasets. A deep auto- comparison between ALO and PSO algorithm was set. The
encoder is utilized in connection with the ALO algorithm. achieved results demonstrated that ALO increases precisely
The effectiveness of the proposed algorithm is empirically while PSO algorithm has a delayed convergence rate. Using
examined and vied for comparison against other published the introduced method, adjusting the design parameters Q
methods in a similar nature utilizing an off-line hand-written and R in UKF detective becomes a simple job with a high
digits picture recognition. The obtained results are compared rate of precision and the restrictions of time cost are over-
with previous methods. Moreover, the effectiveness and come. Also, the introduced method (ALO-UKF) is a useful
applicability of the proposed framework are investigated. way to enhance the predicting effectiveness of requirements
Although the proposed framework with ALO needs some- and coefficients’ difficulties of the selection motor. Simi-
what longer running times, which are a decent trade-off lar method can be used to errors monitoring by an online
when precision is of interest in real-world problems, it is description of their similar signatures.
necessary to additional recognize such frameworks.
4.6 Networks Applications

4.5 Software Engineering In [28], a new approach is suggested using ALO algorithm in

a radial distribution system to voltage profile enhancement
The object-oriented testing paradigm is the most universally and recognizing the optimal system reconfiguration to mini-
employed design paradigm in the information technology mize the energy loss. In network reconfiguration procedures,

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1411

the construction of distribution in supplying material to a based on a decrease of the whole discovering failures and
machine (feeders) is adjusted by adjusting the open or closed determine the near-optimal outlining of thyristor established
situation for the section utilization and tie-line switches in series reparations. The proposed approach is verified using
a spiral configuration mode. Through reconfiguring the net- various regular distribution examination methods IEEE-33-
work operation, the real energy falls could be decreased and bus standard simulated results of this approach has been
simultaneously the voltage status can also be increased. The compared to Bird Swarm Algorithm (BSA), Genetic algo-
introduced model is performed on the IEEE sixty nine-bus rithm (GA), and other optimization algorithms. The obtained
distribution standard examination system. The simulation results revealed that the ALO is adequate algorithm to
implemented in different loading condition. The results show achieve better results in regards to reduce power loss.
that ALO algorithm is capable to reduce energy losses and
enhance the energy status at different loading situation.
The problem of community discovery in complicated 5 Discussion
environments has confirmed an extended volume of atten-
tion since the earlier decade. Community discovery is a This section discusses theoretical aspects, assessment, and
method to identify the building components of the network evaluation of Ant Lion Optimizer algorithm.
by grouping the nodes into groups (communities). This pro- Global numerical optimization problems like benchmark
cedure produced the communities for including the intensive functions are linked with a method of achieving near-opti-
connection between the nodes than the community’s inside mal solutions of a mathematical model through finding the
connections. In this paper [101], propose to implement desired minimum or maximum function. Because of the
modified ALO algorithm for effectively discovering the improving the complication of optimization jobs utilized in
communities in wide networks. The introduced method opti- several optimization problems, the improvement of effective
mizes the fitness function and it is capable to identify density stochastic methods is becoming much important, critical,
joined groups of nodes owning scattered interconnects. This and necessary than before. Over the past decennium, several
approach is examined on Bottlenose Dolphins, Zachary’s different optimization methods have been employed accord-
Karate Club, American college football and Books about ing to various aspects of biology or nature-life.
US politics network benchmarks. The obtained results are Generating the candidate solutions, these optimization
matched with the Enhanced Firefly algorithm (EFF) and algorithms are classified into a couple of subclasses of
the Ant Colony Optimization (ACO) algorithms. The result individual-based optimization techniques (one solution)
shows that the introduced algorithm has superior ability than and population-based optimization techniques (solutions).
EFF and ACO for Books about US politics and Zachary, and Large gradient data is required by the above mentioned
provides better outcomes than EFF for American Football that is individual-based optimization algorithms and it
Club and ACO for Dolphins. The experimental results rep- usually; this type works via one single unsystematic solu-
resent the potential of the approach to efficiently classify the tion, improved over a specified number of iterations. The
network structure. less mathematical calculation is needed by these kinds of
In [102], ALO algorithm also is proposed to solve the algorithms and a good approach is implemented to obtain
power problem (voltage) in spiral configuration networks. the near-optimal solution (globally) for the sake of easy
The proposed approach guarantees system operation within patterns or problems. But, they normally have defects such
power standard boundaries and with least power losses. The as illation mechanism, early convergence (fast), and sen-
introduced approach proposes these jobs; voltage stability sitive search (unstable) in complex situations. Concern-
index calculation and load flow analysis with forwarding/ ing the last, the stochastic procedures trend to start with
backward sweep method, optimizing the distributed gen- a population of stochastic solutions, generated from the
erators (size, location, and power factor) by employing the available search regions and developed by the iterations,
ALO. The effectiveness of combined these jobs together one by one. It is a very important technique that utilized
will be capable to calculate the voltage stability index; give this kind of generating solutions to avert trapped in local
a general overview of voltage stability for radial distribu- search. Furthermore, the mutual information across the
tion system; identify the more susceptible node to fail and candidate solutions further supports the used technique
gives a solution (to decrease voltage sensitivity problem grows although several dilemmas of complex-wide search
and decrease system energy losses). The proposed method regions [104].
implemented through the simulation case on IEEE sixty The distribution of the addressing rule is classified into
nine-bus standard and the result gained the optimal solution. a couple of stages, which is the prime advantage of multi-
In [103] propose to use the ALO algorithm in the radial solutions methods (i.e., stochastic search: exploration and
distribution grids. The main goal of this approach to dis- exploitation) [105]. The above-mentioned (exploration)
tributed generations (DGs) in radial distribution networks indicates a manner where the population (solutions) trend

13
1412 L. Abualigah et al.

to be developed repeatedly and examine the encouraging These optimization methods are hybridization and elitism:
areas of the search space as wide as possible. Dissimilarity, the existence of an elite as a dominating element in a system
the search solutions are controlled or led by the exploitation [106, 107].
or intensification to draw close the near optimal solution ALO algorithm has grown exponentially and become
obtained in the global search stage. a powerful mechanism for tackling complex problem, as
The converging on local search has resulted in a fast con- other optimization techniques. It is an artificial intelligent
vergence, but, the powerful local search strategy is escaped technique, works by stochastic computational mechanisms,
from this problem and the result id maintaining the solutions to determine the near-optimal solution for the multi-dimen-
diversification. Therefore, a powerful procedure is able to sional functions and one-dimensional according to its objec-
work in producing a normal equipoise among these search tive function. The ALO algorithm runs in a common pro-
methods (i.e., exploration and exploitation). More newly, cedure, it is easily implemented and easily used in a wide
meta-heuristic optimization methods have worked notably variety of domains. The results summarized in this review
in addressing a widespread variety of real-world optimiza- paper give evidence and proof of performance achievement
tion problems because of its features such as adaptability, of the ALO algorithm in regards of effectiveness and accu-
naturalness, privation-free technicality, and escape the local racy of the obtained best result. This assurance is determined
optima. For this reason, these methods (the population- according to exploring and analyzing the obtained results
based optimization algorithms) employed to solve various among the ALO algorithm and recent published optimiza-
problems quickly. Researchers applied several strong and tion techniques.
valuable optimization approaches because it is motivated Alike to several optimization techniques, ALO algorithm
by nature-life behaviors. owns some benefits (strong points) and few disadvantages
The benefit of meta-heuristic algorithms is illustrated for (weaknesses). In spite of the fact that there is no concourse
many causes. First cause, the utility of random procedure evidence for this optimizer, the final characterizations are
supports these algorithms to avoid the stuck in local search summarized in this review to prove that the ALO competi-
and converge to the near-optimal solution. Presently, the tiveness over other optimization algorithms in regards to
goal is not to get the best solution of the problem solutions convergence accelerated (rate). Table 3 reviews the strength
but to determine the near-optimal solution (with high-grade (advantages) and weaknesses (disadvantages) of ALO
fineness within a sensible running time). The main factor in algorithm.
producing this objective is to get the proper balance across The main challenge of ALO algorithm is how to deter-
search strategies. Exploration propose is to determine the mine the probabilistic convergence characteristics of ALO,
more encouraging areas in a complex-wide search area. Con- which is needed to fully grasp the given technique. The issue
sequently, the exploitation strategy is enhanced (intensified) of early convergence (premature convergence) in the ALO
by the local search strategy in an encouraging area to pro- usually begins the improvement procedure to be stuck dur-
duce better solutions. The optimal execution and effective- ing the exploitation search. This issue usually happens while
ness of a particular optimization technique in trade-off these the solutions divergence reductions and the solutions cannot
methods, the preferable ability will be achieved. The current avoid falling in the local optima. Moreover, there are big
meta-heuristic algorithms achieve the balance among these potentials for researchers to utilize and apply the advantages
two methods In different ways. They could be further modi- of ALO algorithm to address the involved industry and any
fied and adjusted for local or for global search strategies. real-word problems.
Two or more optimization techniques are merged together Based on the previous studies, there are developing mod-
by hybrid its components to get its strong features together els in general direction appeared of the year 2015 to 2018.
and powerful than each one alone while averting as much as Figure 1 presents different kind of hybridization, modifica-
potential their disadvantages and shortcomings. The second tions, binary, alternatives, and applications in regards to the
cause, the success of optimization techniques is due to its whole published papers per year. As observed in the figure,
principle, abilities, simplicity and adjust experimentally. 2018 presents the highest use of ALO algorithm compared
Meta-heuristic optimization algorithms have been utilized to the rest years. This certainly proved that the utilize of
widely to address different problems. However, for complex the ALO algorithm in that domain has gain attention and
cases, the largest part of the optimization techniques still benefit after one year. Figures 2 and 3 demonstrated clearly
yield trapped in local search and be unsuccessful to achieve that the ALO was employed to various different optimization
the near-global solution. This is the purpose of the week of methods and problems.
diversification (global search) part (i.e., component) in the
used technique. Several diversification search strategies are
employed to enhance the effectiveness of well-known opti-
mization techniques and assist in preventing the drawbacks.

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1413

Table 3  Advantages and disadvantages of ALO algorithm

Advantages
Simplicity of hybridisation with other algorithms
Fast convergence speed
Accelerated process in finding excellent solutions
Suitable for many types of optimization problems
An efficient global procedure to search
Suitable for wide search space (continuous and discrete)
Powerful neighborhood search characteristic
Adaptability, robustness, and scalability are found important characteristics
Strong in controlling a wide number of decisions
Have higher feasibility and efficiency in getting global optima
Lower reasonably of stuck in local optima
Less reliance on initial solutions
Simple implementation
Reasonable execution time
A small number of tuning parameters
Disadvantages
The first proposal of ALO has been introduced for continuous optimization problems
Suffer from premature convergence
No theoretical converging characteristic
Likelihood distribution changes by generations

6 Conclusion and Possible Future Directions For the possible future directions, we suggest to utilize
and enhance the ALO algorithm with other techniques
A comprehensive literature review of the ALO algorithm is components for further improvements to solve various
given in this paper. The main objective of this paper is aimed optimization problems. In future work, we will focus on
at summarizing the variants and the application of ALO the following viewpoints:
algorithm in all classifications reviewed. According to the
papers published in the literature, the improved ALO algo- • Adapting the ALO algorithm to solve the unsolved
rithm is the main procedure used. We additionally collected optimization problems by ALO algorithm
all the given details from the papers that applied the ALO • Utilizing the ALO algorithm to solve multi-objective
algorithm such as in benchmark functions, machine learning, optimization problems.
engineering, image processing, software engineering, and • Modifying the ALO algorithm to deal with real-world
networks and other problems. Besides that, many research- problems such as NP-hard problems and discrete opti-
ers observed that ALO algorithm could solve other differ- mization problems.
ent optimization problems such as clustering optimization • Hybridizing the ALO algorithm with other meta-heu-
problems, pattern recognition problems, and unconstrained ristic components to deal with real-world problems
optimization problems. In one word, ALO algorithm is very such as NP-hard problems and discrete optimization
active in all the examination problems based on the review problems.
the published papers in the literature.
As the conclusion, there is still opportunity or space
for improving the performance, the ALO algorithm could
be extended into different hybridization, modification,
amended, and variants based on the needs of the given Compliance with ethical standards
problems. Consequently, the results of this review paper
could be touched by the interested researchers to show Conflict of interest The authors declare that they have no conflict of
interest.
several procedures for enhancement goal by taking into
consideration on the applications, advantages or defects
of other introduced methods by researchers.

13
1414 L. Abualigah et al.

References 23. Mani M, Bozorg-Haddad O, Chu X (2018) Ant lion optimizer

(ALO) algorithm. In: Advanced optimization by nature-
inspired algorithms. Springer, pp 105–116
1. Abualigah L (2020) Multi-verse optimizer algorithm: a com-
24. Tung NS, Chakravorty S (2016) Ant lion optimizer based
prehensive survey of its results, variants, and applications. Neur
approach for optimal scheduling of thermal units for small
Comput Appl 1–21
scale electrical economic power dispatch problem. Int J Grid
2. Bolaji AL, Al-Betar MA, Awadallah MA, Khader AT, Abualigah
Distrib Comput 9:211–224
LM (2016) A comprehensive review: Krill herd algorithm (kh)
25. Abualigah LMQ (2019) Feature selection and enhanced krill
and its applications. Appl Soft Comput 49:437–446
herd algorithm for text document clustering. Springer, Berlin
3. Glover F (1977) Heuristics for integer programming using sur-
26. Mafarja MM, Mirjalili S (2018) Hybrid binary ant lion opti-
rogate constraints. Decis Sci 8:156–166
mizer with rough set and approximate entropy reducts for fea-
4. Kirkpatrick S, Gelatt CD, Vecchi MP (1983) Optimization by
ture selection. Soft Comput 1–17
simulated annealing. Science 220:671–680
27. Wu Z, Yu D, Kang X (2017) Parameter identification of pho-
5. Koziel S, Yang X-S (2011) Computational optimization, methods
tovoltaic cell model based on improved ant lion optimizer.
and algorithms, vol 356. Springer, Berlin
Energy Convers Manag 151:107–115
6. Abualigah LM, Sawaie AM, Khader AT, Rashaideh H, Al-Betar
28. Zainal MI, Yasin ZM, Zakaria Z (2017) Network reconfigura-
MA, Shehab M (2017) 𝛽 -hill climbing technique for the text
tion for loss minimization and voltage profile improvement
document clustering. New Trends Inf Technol (NTIT) 60
using ant lion optimizer. In: IEEE Conference on Systems,
7. Shehab M, Abualigah L, Al Hamad H, Alabool H, Alshinwan
Process and Control (ICSPC), IEEE, pp 162–167
M, Khasawneh AM (2019) Moth–flame optimization algorithm:
29. Nair SS, Rana K, Kumar V, Chawla A (2017) Efficient mod-
variants and applications. Neur Comput Appl 1–26
eling of linear discrete filters using ant lion optimizer. Circuits
8. Abualigah L, Shehab M, Alshinwan M, Alabool H (2019) Salp
Syst Signal Process 36:1535–1568
swarm algorithm: a comprehensive survey. Neur Comput Appl
30. Mouassa S, Bouktir T, Salhi A (2017) Ant lion optimizer for
1–21
solving optimal reactive power dispatch problem in power sys-
9. Abdelmadjid C, Mohamed S-A, Boussad B (2013) Cfd analysis
tems. Eng Sci Technol Int J 20:885–895
of the volute geometry effect on the turbulent air flow through
31. Dinkar SK, Deep K (2017) Opposition based laplacian ant lion
the turbocharger compressor. Energy Procedia 36:746–755
optimizer. J Comput Sci 23:71–90
10. Mirjalili S, Gandomi AH, Mirjalili SZ, Saremi S, Faris H, Mir-
32. Rayyam M, Zazi M, Barradi Y (2018) A new metaheuristic
jalili SM (2017) Salp swarm algorithm: a bio-inspired optimizer
unscented kalman filter for state vector estimation of the induc-
for engineering design problems. Adv Eng Softw 114:163–191
tion motor based on ant lion optimizer. COMPEL-Int J Comput
11. Kennedy J (2011) Particle swarm optimization. In: Encyclopedia
Math Electr Electr Eng 37:1054–1068
of machine learning, Springer, pp 760–766
33. Eltag K, Aslamx MS, Ullah R (2019) Dynamic stability
12. Abualigah LM, Khader AT (2017) Unsupervised text feature
enhancement using fuzzy pid control technology for power
selection technique based on hybrid particle swarm optimiza-
system. Int J Control Autom Syst 17:234–242
tion algorithm with genetic operators for the text clustering. J
34. Wang M, Wu C, Wang L, Xiang D, Huang X (2019) A feature
Supercomput 73:4773–4795
selection approach for hyperspectral image based on modified
13. Mirjalili S, Mirjalili SM, Lewis A (2014) Grey wolf optimizer.
ant lion optimizer. Knowl-Based Syst 168:39–48
Adv Eng Softw 69:46–61
35. Gupta S, Kumar V, Rana K, Mishra P, Kumar J (2016) Devel-
14. Rashedi E, Nezamabadi-Pour H, Saryazdi S (2009) Gsa: a gravi-
opment of ant lion optimizer toolkit in labviewTM . In: Inter-
tational search algorithm. Inf Sci 179:2232–2248
national Conference on Innovation and Challenges in Cyber
15. Mirjalili S (2015) Moth-flame optimization algorithm: a
Security (ICICCS-INBUSH), IEEE, pp 251–256
novel nature-inspired heuristic paradigm. Knowl-Based Syst
36. Farughi H, Mostafayi S, Arkat J (2019) Healthcare districting
89:228–249
optimization using gray wolf optimizer and ant lion optimizer
16. Yang X-S (2010) A new metaheuristic bat-inspired algorithm. In:
algorithms (case study: South khorasan healthcare system in
Nature inspired cooperative strategies for optimization (NICSO
Iran). J Optim Ind Eng 12:119–131
2010), Springer, pp 65–74
37. Gutiérrez JL, Rivera SR (2018) Benchmark functions opti-
17. Karaboga D, Basturk B (2007) A powerful and efficient algo-
mization using binary biogeography-based optimization with
rithm for numerical function optimization: artificial bee colony
aleatory-mixed migration (BBBO-AMM) and binary ant-lion
(abc) algorithm. J Global Optim 39:459–471
optimizer (BALO)
18. Abualigah LMQ (2019) Krill herd algorithm. In: Feature selec-
38. Bayati Chaleshtari MH, Ma Jafari (2017) Optimization of finite
tion and enhanced Krill Herd algorithm for text document clus-
metallic plates with quadrilateral cutout subjected to in- plane
tering. Springer, pp 11–19
loading by ant lion optimizer. Modares Mech Eng 17:11–22
19. Abualigah LM, Khader AT, Hanandeh ES (2019) Modified krill
39. Yao P, Wang H (2017) Dynamic adaptive ant lion optimizer
herd algorithm for global numerical optimization problems.
applied to route planning for unmanned aerial vehicle. Soft
In: Advances in nature-inspired computing and applications.
Comput 21:5475–5488
Springer, pp 205–221
40. Ksiazek K, Polap D, WoZniak M, Damavsevicius R (2017)
20. Yang X-S, Deb S (2009) Cuckoo search via lévy flights. In: World
Radiation heat transfer optimization by the use of modified ant
congress on nature and biologically inspired computing (NaBIC).
lion optimizer. In: IEEE Symposium Series on Computational
IEEE, pp 210–214
Intelligence (SSCI). IEEE, pp 1–7
21. Mirjalili S (2015) The ant lion optimizer. Adv Eng Softw
41. Rajan A, Jeevan K, Malakar T (2017) Weighted elitism based
83:80–98
ant lion optimizer to solve optimum var planning problem. Appl
22. Abualigah L, Diabat A (2020) A novel hybrid antlion optimiza-
Soft Comput 55:352–370
tion algorithm for multi-objective task scheduling problems in
42. Radha J, Subramanian S, Ganesan S, Abirami M (2017) Emis-
cloud computing environments. Clust Comput 1–19
sion/fuel energy restricted dynamic optimal power flow using
fuzzy-ant lion optimizer. Int J Energy Sect Manag 11:208–256

13
Ant Lion Optimizer: A Comprehensive Survey of Its Variants and Applications 1415

43. Zhao X, Jing K, Liu D, Yan X (2018) Improved ant lion optimizer 61. Umamaheswari E, Ganesan S, Abirami M, Subramanian S
and its application in modeling of czochralski crystal growth. In: (2016) Deterministic reliability model based preventive genera-
Chinese Control And Decision Conference (CCDC). IEEE, pp tor maintenance scheduling using ant lion optimizer. In: Interna-
3106–3113 tional conference on circuit, power and computing technologies
44. Jiang F, He J, Peng Z (2018) Short-term wind power forecasting (ICCPCT), IEEE, pp 1–8
based on bp neural network with improved ant lion optimizer. In: 62. Umamaheswari E, Ganesan S, Abirami M, Subramanian S
37th Chinese Control Conference (CCC). IEEE, pp 8543–8547 (2017a) Stochastic model based reliability centered preventive
45. Yang D, Miao J, Zhang F, Tao J, Wang G, Shen Y (2019) Bearing generator maintenance planning using ant lion optimizer. In:
fault diagnosis using a support vector machine optimized by an International conference on circuit, power and computing tech-
improved ant lion optimizer. Shock Vibr 2019 nologies (ICCPCT), IEEE, pp 1–8
46. Guo S, Zhao H, Zhao H (2017) A new hybrid wind power fore- 63. Umamaheswari E, Ganesan S, Abirami M, Subramanian
caster using the beveridge-nelson decomposition method and a S (2017b) Cost effective integrated maintenance schedul-
relevance vector machine optimized by the ant lion optimizer. ing in power systems using ant lion optimizer. Energy Proc
Energies 10:922 117:501–508
47. Majhi SK, Biswal S (2018) Optimal cluster analysis using hybrid 64. Silberschatz A, Gagne G, Galvin PB (2018) Operating system
k-means and ant lion optimizer. Karbala Int J Mod Sci 4:347–360 concepts. Wiley, New York
48. Fathy A, Abdelaziz AY (2018) Single and multi-objective oper- 65. Elango U, Sivarajan G, Manoharan A, Srikrishna S (2018) Pre-
ation management of micro-grid using krill herd optimization ventive maintenance scheduling using analysis of variance-based
and ant lion optimizer algorithms. Int J Energy Environ Eng ant lion optimizer. World J Eng 15:254–272
9:257–271 66. Utama DM, Baroto T, Maharani D, Jannah FR, Octaria RA
49. Wang M, Gao L, Huang X, Jiang Y, Gao X (2019) A texture (2019) Algoritma ant-lion optimizer untuk meminimasi emisi
classification approach based on the integrated optimization for karbon pada penjadwalan flow shop dependent sequence set-up.
parameters and features of gabor filter via hybrid ant lion opti- Jurnal Litbang Industri 9:69–78
mizer. Appl Sci 9:2173 67. Dinkar SK, Deep K (2019) A novel cpu scheduling algorithm
50. Azizi M, Ghasemi SAM, Ejlali RG, Talatahari S (2019) Optimum based on ant lion optimizer. In: Soft computing for problem solv-
design of fuzzy controller using hybrid ant lion optimizer and ing, Springer, pp 339–353
jaya algorithm. Artif Intell Rev 1–32 68. Trivedi I, Jangir P, Parmar SA, Motilal B, Jangir N, Kumar A
51. Zhao S, Gao L, Yu D, Tu J (2016) Ant lion optimizer with chaotic (2016) Power system engineering optimization using ant lion
investigation mechanism for optimizing SVM parameters. J Front optimizer
Comput Sci Technol 10:722–731 69. Engel EA, Kovalev IV (2016) MPPT of a partially shaded photo-
52. Kose U (2018) An ant-lion optimizer-trained artificial neural net- voltaic module by ant lion optimizer. In: International conference
work system for chaotic electroencephalogram (EEG) prediction. on swarm intelligence. Springer, pp 451–457
Appl Sci 8:1613 70. Trivedi IN, Parmar SA, Bhesdadiya R, Jangir P (2016) Voltage
53. Mirjalili S, Jangir P, Saremi S (2017) Multi-objective ant lion stability enhancement and voltage deviation minimization using
optimizer: a multi-objective optimization algorithm for solving ant-lion optimizer algorithm, In: 2nd international conference on
engineering problems. Appl Intell 46:79–95 advances in electrical, electronics, information, communication
54. Hosseini K, Araghi S, Ahmadian MB, Asadian V (2017) Multi- and bio-informatics (AEEICB). IEEE, pp 263–267
objective optimal scheduling of a micro-grid consisted of renew- 71. Parmar S, Trivedi I, Bhesdadiya R, Jangir P (2016) Reactive and
able energies using multi-objective ant lion optimizer. In: Smart active power loss reduction using ant-lion optimizer
Grid Conference (SGC). IEEE, 1–8 72. Yasin Z, Mohamad H, Wahab N, Sam’on I (2017) Optimal under-
55. Abul’Wafa AR (2019) Ant-lion optimizer-based multi-objective voltage load shedding using ant lion optimizer. Int J Simul: Syst
optimal simultaneous allocation of distributed generations and Sci Technol 10:1–6
synchronous condensers in distribution networks. Int Trans 73. Mansour HS, Abdelsalam AA, Sallam AA (2017) Optimal dis-
Electr Energy Syst 29:e2755 tributed energy resources allocation using ant-lion optimizer for
56. Mousavifard R, Abolghasemzadeh M, Razmjooy N, Alizadeh Y power losses reduction. In: IEEE international conference on
(2019) Optimal design of functionally graded steels using multi- smart energy grid engineering (SEGE). IEEE, pp 346–352
objective ant lion optimizer 74. Hatata AY, Hafez AA (2018) Ant lion optimizer versus particle
57. Dhifaoui C, Kahouli O, Abdallah HH (2019) Multi-objective swarm and artificial immune system for economical and eco-
ant lion optimizer to solve the dynamic economic dispatch prob- friendly power system operation. Int Trans Electr Energy Syst
lem with valve point effect. In: 19th International conference e2803
on sciences and techniques of automatic control and computer 75. Špoljarić T, Pavić I (2018) Performance analysis of an ant lion
engineering (STA), IEEE, pp 564–571 optimizer in tuning generators’ excitation controls in multi
58. Wang R-A, Zhou Y-W, Zheng Y-Y (2018) Ant lion optimizer machine power system. In: 41st International convention on
with adaptive boundary and optimal guidance. In: International information and communication technology, electronics and
conference on mechatronics and intelligent robotics. Springer, microelectronics (MIPRO). IEEE, pp 1040–1045
pp 379–386 76. Costa Filho RND, Paucar VL (2018) Robust and coordinated
59. Hu P, Wang Y, Wang H, Zhao R, Yuan C, Zheng Y, Lu Q, Li tuning of pss and facts-pods of interconnected systems consider-
Y, Masood I (2018) Alo-dm: A smart approach based on ant ing signal transmission delay using ant lion optimizer. J Control
lion optimizer with differential mutation operator in big data Autom Electr Syst 29:625–639
analytics. In: International conference on database systems for 77. Hatata A, Lafi A (2018) Ant lion optimizer for optimal coordina-
advanced applications, Springer, pp 64–73 tion of doc relays in distribution systems containing DGS. IEEE
60. Heidari AA, Faris H, Mirjalili S, Aljarah I, Mafarja M (2020) Ant Access 6:72241–72252
lion optimizer: theory, literature review, and application in multi- 78. Ali AH, Youssef A-R, George T, Kamel S (2018) Optimal dg
layer perceptron neural networks. In: Nature-inspired optimizers, allocation in distribution systems using ant lion optimizer. In:
Springer, pp 23–46

13
1416 L. Abualigah et al.

International conference on innovative trends in computer engi- 93. Kanimozhi G, Kumar H (2018) Modeling of solar cell under dif-
neering (ITCE), IEEE, pp 324–331 ferent conditions by ant lion optimizer with lambertw function.
79. Mei R, Sulaiman M, Mustaffa Z (2015) Ant lion optimizer for Appl Soft Comput 71:141–151
optimal reactive power dispatch solution. J Electr Syst Spec Issue 94. Srinivasan K, Soundirarrajan N (2019) Performance evaluation
AMPE 2016:68–74 of ant lion optimizer-based PID controller for speed control of
80. Balachandar P, Ganesan S, Jayakumar N, Subramanian S (2017) PMSM. J Test Eval 49
Multi-fuel power dispatch in an interconnected power system 95. Talatahari S (2016) Optimum design of skeletal structures using
using ant lion optimizer: multi-fuel dispatch considering tie-line ant lion optimizer. Iran Univ Sci Technol 6:13–25
limits. Int J Energy Optim Eng (IJEOE) 6:29–54 96. Chaleshtari MHB, Jafari M (2019) Ant lion optimizer for opti-
81. Kamboj VK, Bhadoria A, Bath S (2017) Solution of non-convex mization of finite perforated metallic plate. Struct Eng Mech
economic load dispatch problem for small-scale power systems 69:667–676
using ant lion optimizer. Neur Comput Appl 28:2181–2192 97. Sam’on IN, Yasin ZM, Zakaria Z (2017) Ant lion optimizer for
82. Mei RNS, Sulaiman MH, Daniyal H, Mustaffa Z (2018) Appli- solving unit commitment problem in smart grid system. Indones
cation of moth-flame optimizer and ant lion optimizer to solve J Electr Eng Comput Sci 8:129–136
optimal reactive power dispatch problems, Journal of Telecom- 98. Roy K, Mandal KK, Mandal AC (2019) Ant-lion optimizer algo-
munication. Electron Comput Eng (JTEC) 10:105–110 rithm and recurrent neural network for energy management of
83. Júnior JdAB, Nunes MVA, Nascimento MHR, de Freitas CAO, micro grid connected system. Energy 167:402–416
Leite JC, Moraes NM (2018a) Ant lion optimizer applied to eco- 99. Leke CA, Marwala T (2019) Missing data estimation using ant-
nomic emission load dispatch problems turning off the engines. lion optimizer algorithm. In: Deep learning and missing data in
In: 13th IEEE international conference on industry applications engineering systems. Springer, pp 103–114
(INDUSCON). IEEE, pp 829–836 100. Sharma R, Saha A (2019) Ant lion optimizer for state based
84. Júnior JdAB, Nascimento MHR, De Freitas CAO, Leite JC, object oriented testing. J Inf Optim Sci 40:219–232
Carvajal TLR (2018b) Approach of economic-emission load 101. Kaur M, Mahajan A (2017) Community detection in complex
dispatch using ant lion optimizer. Int J Adv Eng Res Sci 5 networks: a novel approach based on ant lion optimizer. In: Pro-
85. Alazemi FZ, Hatata AY (2019) Ant lion optimizer for optimum ceedings of 6th international conference on soft computing for
economic dispatch considering demand response as a visual problem solving. Springer, pp 22–34
power plant. Electr Power Compon Syst 1–15 102. Maher M, Ebrahim M, Mohamed E, Mohamed A (2017) Ant-lion
86. Hatata AY, Hafez AA (2019) Ant lion optimizer versus particle optimizer based optimal allocation of distributed generators in
swarm and artificial immune system for economical and eco- radial distribution networks. Int J Eng Inf Syst 1:225–238
friendly power system operation. Int Trans Electrical Energy Syst 103. George T, Youssef A-R, Kamel S (2018) Optimal allocation
29:e2803 of DGS and TCSC in radial networks using ant lion optimizer.
87. Saikia LC, Sinha N (2016) Automatic generation control of a In: 20th International middle east power systems conference
multi-area system using ant lion optimizer algorithm based pid (MEPCON). IEEE, pp 1092–1097
plus second order derivative controller. Int J Electr Power Energy 104. Gharehchopogh FS, Gholizadeh H (2019) A comprehensive sur-
Syst 80:52–63 vey: Whale optimization algorithm and its applications. Swarm
88. Prasad ES, Ram BS (2016) Ant-lion optimizer algorithm based and evolutionary computation
FOPID controller for speed control and torque ripple minimiza- 105. Malhotra R, Khanna M, Raje RR (2017) On the application of
tion of srm drive system. In: International conference on sig- search-based techniques for software engineering predictive
nal processing, communication, power and embedded system modeling: a systematic review and future directions. Swarm Evol
(SCOPES). IEEE, pp 1550–1557 Comput 32:85–109
89. Pradhan R, Majhi SK, Pradhan JK, Pati BB (2017) Performance 106. Rakshit P, Konar A, Das S (2017) Noisy evolutionary optimiza-
evaluation of pid controller for an automobile cruise control sys- tion algorithms-a comprehensive survey. Swarm Evol Comput
tem using ant lion optimizer. Eng J 21:347–361 33:18–45
90. Mokeddem D, Draidi H (2018) Optimization of PID sliding sur- 107. Gotmare A, Bhattacharjee SS, Patidar R, George NV (2017)
face using ant lion optimizer. In: International symposium on Swarm and evolutionary computing algorithms for system iden-
modelling and implementation of complex systems. Springer, tification and filter design: A comprehensive review. Swarm Evol
pp 133–145 Comput 32:68–84
91. Spoljarić T, Lušetić C, Simovic V (2018) Optimization of pid
controller in AVR system by using ant lion optimizer algorithm. Publisher’s Note Springer Nature remains neutral with regard to
In: 41st International convention on information and communi- jurisdictional claims in published maps and institutional affiliations.
cation technology, electronics and microelectronics (MIPRO).
IEEE, pp 1522–1526
92. Pradhan R, Majhi SK, Pati BB (2018) Design of pid controller
for automatic voltage regulator system using ant lion optimizer.
World J Eng 15:373–387

13