See discussions, stats, and author profiles for this publication at: [Link]

net/publication/376582239

Robot Path Planning Using Grasshopper Optimization Algorithm

Conference Paper · December 2023

CITATIONS READS

0 218

2 authors:

Tohid Yousefi Özlem Varlıklar

Dokuz Eylül University Dokuz Eylül University
11 PUBLICATIONS 13 CITATIONS 43 PUBLICATIONS 350 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Tohid Yousefi on 16 December 2023.

The user has requested enhancement of the downloaded file.

 2nd International Conference on Frontiers
in Academic Research
December 4-5, 2023 : Konya, Turkey

AS-Proceedings © 2023 Published by AS-Proceedings

[Link] [Link]

Robot Path Planning Using Grasshopper Optimization Algorithm

Tohid Yousefi*, Özlem Aktaş 2
1
Computer Engineering, Graduate School of Natural and Applied Sciences, Dokuz Eylül University, Turkey
2
Computer Engineering Department, Engineering Faculty, Dokuz Eylül University, Turkey
*
([Link]@[Link]) Email of the corresponding author

Abstract – In the world of mobile robots, figuring out the best path is a big deal. Robots use specific plans
to move from one point to another. The main aim of path planning is to find safe moves for the robot in
places full of obstacles. These moves create a smooth, collision-free path from where the robot starts to its
target. There are many ways to solve this path problem, but they're not perfect yet. Recently, researchers
have turned to smart methods, called meta-heuristic methods, to solve these problems. In our study, we
tried out the particle swarm optimization algorithm (PSO) and the grasshopper optimization algorithm
(GOA). Looking at the results, we found that PSO had a best cost value of 7.5558, and GOA had a best
cost value of 7.5421. This shows how good the grasshopper optimization algorithm is at path planning
and how effective these meta-heuristic methods can be.

Keywords – Grasshopper Optimization Algorithm, Meta-heuristic algorithms, Optimization, Particle Swam Optimization,
Robot Path Planning

I. INTRODUCTION in environments with fixed obstacles. The primary

Robotics is a well-known field widely used in goal is to find paths that avoid collisions for the
industry, and gradually becoming a part of our robot. The robot's objective is to reach the target
daily lives. However, there are still several position swiftly, minimizing the distance between
challenges that need solutions or enhancements. the starting and target positions. This path planning
One of the most important challenges is robot path problem involves defining the geometry of the
planning, which gets harder as robots become more area, specifying the robot's starting and target
independent and our expectations for them grow. points, identifying obstacle locations, and
Nowadays, automatic robots play a very establishing region boundaries [2, 3].
important role in the field of logistics. Various Over time, several path planning algorithms and
types of robots have been created and proven techniques have been introduced and researched to
effective in tasks like managing smart homes, address this challenge. Classical algorithms such as
airports, shopping malls, and manufacturing labs. artificial potential fields, the A* algorithm, and the
For a robot to perform tasks optimally, intelligence Dijkstra algorithm have been commonly utilized.
needs to be integrated. This poses challenges, However, contemporary methods offer more
especially in areas like navigation for mobile practical and improved outcomes [4].
robots. To successfully navigate, a robot must This study aims to employ meta-heuristic
know its position in relation to its goal, consider algorithms to tackle the robot path planning
the surrounding dangers, and adjust actions for the problem. Meta-heuristic algorithms have gained
best chance of reaching the destination [1]. popularity in recent years for optimization
In recent years, the issue of planning the path for problems. These algorithms draw inspiration from
robots has gained significant attention, with evolutionary concepts, physical laws, or the social
various approaches proposed for navigating robots behaviors observed in natural animal herds [5].
105
 In our research, we contributed by utilizing two robot to compensate for the demerits of
specialized optimization algorithms, namely the conventional methods by learning the situation
particle swarm optimization algorithm [6] and the where each robot has a mutual influence. In most
grasshopper optimization algorithm [7]. Our focus cases, existing path-planning algorithms highly
centered on conducting a comprehensive depend on the environment. Although the proposed
comparative analysis of these algorithms within the algorithm used an A* algorithm that could not be
domain of robot path planning. This comparative used in a dynamic environment as a comparison
exploration provides valuable insights into their algorithm, it also showed that path generation is
respective strengths, weaknesses, and applicability possible even in a dynamic environment.
in addressing the challenges associated with In the proposed method [12], an improved
efficient robot path planning. velocity updating algorithm for the particle of PSO
is proposed to improve the convergent speed of the
II. LITERATURE REVIEW PSO-based method. Furthermore, to deal with the
This article [8] introduces an optimized version local optimum problem and realize the global path
of the heuristic path-searching which is based on planning in an unknown environment, a bio-
automated enterprise logistics application inspired neural network is used to calculate the
scenarios. In the automation of enterprise logistics, fitness for PSO based method.
the distance of the path is not the only This paper [13] presents a global path planning
consideration of the weight of the path searching, method based on a grasshopper optimization
we must adjust the valuation function to adapt to algorithm (GOA) in a known static environment.
the complex practical field. At the same time, this This algorithm is improved using the bias factor to
article introduces a method to avoid conflict based increase the efficiency and improve the resulting
on the time window. And proposed a data structure path. The resulting path from this algorithm is
to adapt to the algorithm in the process of frequent further enhanced using an improved version
insertion and searching operations. multinomial logistic regression algorithm (MLR).
In this paper [9], the authors of the article
propose a graph-based Floyd algorithm and a III. MATERIALS AND METHOD
feature-based algorithm to achieve robot path In this section, information is provided on the
optimization. Specifically, for a set of video clips, robot path planning (RPP), optimization, meta-
they first leverage the dense sampling algorithm to heuristic algorithms, particle swarm optimization
extract spatial-temporal interest points from the algorithm (PSO), and grasshopper optimization
video, which is further used for video algorithm (GOA).
segmentation. Then, each video section is
A. Robot Path Planning (RPP)
represented by the local motion and path
appearance, which will be quantized into a path- The process of finding a path without
related feature vector. Finally, they leverage the encountering obstacles from the beginning to
improved linear discriminative analysis to discover reaching a predefined target point in the shortest
both the global and within-class geometry of path possible time is called robot path planning [14]. In
representation. this book [15], path planning is defined as reaching
In this paper [10], to improve the optimality, the from a starting point to an ending point without
authors of the article propose a Forward Search encountering obstacles around it.
Optimization (FSO) algorithm to shorten the path Path planning is a critical aspect of robotics,
planned by searching-based global algorithms such particularly for control engineers, as it addresses
as Dijkstra, A*, D*, and D* Lite. To improve the key challenges in various fields. It involves guiding
practicability, they develop a Subgoal-based robots through tasks, ranging from basic trajectory
Hybrid Path Planning (SHPP) approach to smooth planning to selecting optimal action sequences for
the path and keep it safe and comfortable from specific objectives. The limitation of not always
obstacles. having advance global environmental information
In this paper [11], the proposed algorithm requires the use of algorithms, allowing path
combines the reinforcement learning algorithm planning to be applied in partially or unknown
with the path planning algorithm of the mobile structured environments. The generation of a

106
suitable trajectory involves a sequence of actions, given conditions by minimizing costs or
navigating the robot from the starting point to the maximizing utility. These conditions, leading to the
target through intermediate states. Decisions in maximum or minimum value of a function, are
path planning algorithms are based on current state referred to as optimizations [19].
information and criteria like the shortest distance Optimization is the process of identifying the
measured using Euclidean distance computation. most optimal solution from a set of possibilities. It
While there may be multiple paths, some situations is both a mathematical assessment indicating the
may lack a feasible route. The optimal path is quality of a solution and a method that models a
ideally the shortest, obstacle-free, and takes the problem as an objective function, systematically
least time to reach the goal, considering motion exploring potential solutions to find the best one
constraints like nonholonomic conditions in [20, 21].
underactuated systems. Path planning proves to be As illustrated in the equation below, the
a versatile tool in robotics, offering solutions to a conventional representation of an optimization
variety of real-world problems [16]. problem typically revolves around minimizing the
Today, we are at the peak of progress in the field objective function [22]:
of robotics, and every day new robots with various 𝑚𝑖𝑛 𝑓(𝑥) ∶ 𝑥
applications enter the labor market. One of these
𝑔𝑗 (𝑥) ≥ 0, 𝑗 = {1,2, … , 𝐽}
new systems is mobile robotic systems. which ∈ 𝑅 𝑛 𝑠. 𝑡 {
allows us to do many logistic tasks automatically. ℎ𝑘 (𝑥) = 0, 𝑘 = {1,2, … , 𝐾}
These types of robots need to know their position
(𝐿) (𝑢)
relative to the target position to successfully 𝑥𝑖 ≤ 𝑥𝑖 ≤ 𝑥𝑖 𝑖 = {1,2, … , 𝑛}
navigate. In addition, they should consider the
dangers and obstacles of the surrounding where 𝑓(𝑥) is the optimized function, 𝑔𝑗 (𝑥) is j
environment and adjust their actions to maximize inequality constraints and ℎ𝑘 (𝑥) represents K
the chances of reaching the destination [17]. equality constraints.
The robot path planning problem is a challenging
optimization task commonly addressed using meta- C. Meta-Heuristic Algorithms
heuristic algorithms. The goal is for the mobile Meta-heuristic algorithms mark a notable
robot to reach its target from the starting point via departure from historical heuristic methods,
the shortest path without encountering obstacles. introducing a more systematic and innovative
The problem involves factors like start and target approach to problem-solving. While traditional
positions, obstacle size, shape, number, and zone heuristics often relied on intuition and accidental
boundaries. The objective function for path discoveries, meta-heuristics demonstrate
planning is crucial in defining the optimization widespread success by drawing inspiration from
criteria. Here is the mathematical expression natural events like biological systems and physical
representing the objective function for planning the processes. This shift represents an evolution in
path of a robot [18]: problem-solving, embracing strategic algorithms
inspired by real-world processes for increased
𝑗 = 𝑚𝑖𝑛 𝑄 (1 + 𝛽𝑉) efficiency and adaptability [5, 23].
Meta-heuristics offer an effective means of
Here, 𝛽 stands for the violation coefficient (fixed finding practical solutions to complex problems
at 100), V represents the violation cost, and Q within a reasonable timeframe through trial and
indicates the overall distance between the starting error. Their primary objective is to identify well-
and target points. practiced solutions within an acceptable time scale
[5]. These methods enable efficient exploration of
B. Optimization solution spaces, striking a balance between
Optimization is a crucial aspect across numerous effectiveness and practical time constraints.
applications in engineering and industry, aiming to
minimize costs, energy consumption, or maximize D. Particle Swarm Optimization
profits, output, performance, and efficiency. The The particle swarm optimization is a
objective is to achieve the best outcomes under computational method introduced by Kennedy and
107
Eberhart in 1995 [6]. Since its inception, various global and local search, enhancing the performance
adaptations have been created to address practical of particle swarm optimization [26].
optimization problems [24]. This algorithm
E. Grasshopper Optimization Algorithm
operates as a swarm-based search, with each entity
referred to as a particle. Employed to tackle Grasshoppers pose a significant threat to
optimization challenges, the method involves agriculture, affecting product quality and causing
iteratively adjusting particle velocities based on damage [28]. Their life cycle involves two stages:
combined information and calculating new nymph and adulthood. The nymph stage reflects
positions in each dimension. These particles small, slow movements akin to optimization
continuously navigate a multidimensional search extraction, while adulthood involves sudden, long
space, aiming to achieve equilibrium, optimal steps resembling exploration in optimization
solutions, or surpass computational constraints. algorithms. The Grasshopper Optimization
Figure 1 illustrates the overall flowchart of the Algorithm, introduced in 2017, mimics
particle swarm optimization algorithm [25]. grasshopper swarming behaviors in natural search
processes [7]. This algorithm, categorized as a
swarming algorithm, is inspired by the collective
movement of grasshoppers [29]. Figure 2 displays
the flowchart illustrating the Grasshopper
Optimization Algorithm [7].

Fig. 1 General flowchart of particle swarm optimization

algorithm
Particle swarm optimization (PSO), aims to
discover the optimum solution by navigating a D-
dimensional search space with individual particles.
Each particle, denoted as the i-th particle,
possesses a position vector 𝑥 𝑖 and a corresponding
velocity vector as 𝑣 𝑖 . The personal best position of
any particle is recorded as 𝑥 𝑖,𝑏𝑒𝑠𝑡 , while the best Fig. 2 General flowchart of grasshopper optimization
solution encountered by the entire swarm is termed algorithm
𝑥 𝑔𝑏𝑒𝑠𝑡 . The initial swarm comprises a population The mathematical model for grasshopper swarm
of randomly generated solutions. The algorithm behavior is represented by below Equation [7]:
iteratively refines the positions and velocities of
𝑛𝑃𝑜𝑝
particles based on 𝑥 𝑖,𝑏𝑒𝑠𝑡 and 𝑥 𝑔𝑏𝑒𝑠𝑡 , employing 𝑢𝑏𝑑 − 𝑙𝑏𝑑
the following formulas [26]: 𝑋𝑖𝑑 (𝑡 + 1) = 𝑐 ∑𝑐 𝑠(|𝑥𝑗𝑑
2
𝑗=1
𝑣 𝑖 [𝑡 + 1] = 𝑤𝑣 𝑖 [𝑡] + 𝑐1 𝑟1 (𝑥 𝑖,𝑏𝑒𝑠𝑡 [𝑡] − 𝑥 𝑖 [𝑡]) ( (𝑗≠𝑖)
+ 𝑐2 𝑟2 (𝑥 𝑔𝑏𝑒𝑠𝑡 [𝑡] − 𝑥 𝑖 [𝑡])
𝑥(𝑗) − 𝑥(𝑖)
− 𝑥𝑖𝑑 | ) + 𝑇̂𝑑
𝑥 𝑖 [𝑡 + 1] = 𝑥 𝑖 [𝑡] + 𝑣 𝑖 [𝑡 + 1] 𝑑𝑖𝑗
)
Where 𝑡 shows that the algorithm is ongoing at In the equation, 𝑢𝑏𝑑 represents the upper bound
the t-th iteration. The constants 𝑐1 and 𝑐2 are in the d dimension, and 𝑙𝑏𝑑 is the lower bound in
accelerations, and 𝑟1 and 𝑟2 are random values the 𝑑 dimension. 𝑇̂𝑑 signifies the value of the 𝑑
ranging from 0 to 1. The inertia weight 𝑤, dimension in the target (the best solution found so
introduced by Shi and Eberhart [27], helps balance far). The parameter 𝑐 is a diminishing coefficient
108
that contracts the comfort zone, repulsion zone, use the PSO algorithm, we must determine the
and attraction zone. The initial 𝑐 is comparable to value of some hyperparameters in advance, which
the inertia weight (ω) in the Particle Swarm you can see in Table 1.
Optimization Algorithm, restraining grasshopper Table 1. The PSO algorithm's hyperparameter values
movements around the target and balancing swarm
exploration and exploitation. The second "c" is c1 c2 w MaxIt nPop
1.5 2 1 1000 500
utilized to decrease the repulsion zone, gravity
zone, and comfort zone between grasshoppers
We used the PSO algorithm, using the settings in
relative to the number of iterations. The calculation
Table 1, to assist the robot in moving past obstacles
for 𝑐 is expressed in the following equation, where
and reaching the goal shown in Figure 3. Here are
it is considered as a single parameter.
the results we obtained.
𝑐𝑚𝑎𝑥 − 𝑐𝑚𝑖𝑛
𝑐 = 𝑐𝑚𝑎𝑥 − 𝑡
𝑡𝑚𝑎𝑥
𝑐𝑚𝑎𝑥 = 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑉𝑎𝑙𝑢𝑒 𝐶(≈ 1)
,
𝑐𝑚𝑖𝑛 = 𝑀𝑖𝑛𝑖𝑚𝑢𝑚 𝑉𝑎𝑙𝑢𝑒 𝐶(≈ 0)
{
𝑡 = 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛
𝑡𝑚𝑎𝑥 = 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐼𝑡𝑒𝑟𝑎𝑡𝑖𝑜𝑛

IV. RESULTS
In this section of our research, we analyze the
outcomes achieved by employing the PSO and
GOA algorithms for resolving the robot path
planning issue.
Figure 3 shows an example of how a robot plans Fig. 4 The best solution obtained by the particle swarm
its path. In a 6x6 area, there are three round optimization algorithm
obstacles with different sizes. The yellow square is
In Figure 4, there's a clear demonstration that the
where the robot starts, and the green star is where it
PSO algorithm successfully navigated through the
wants to go. We use PSO and GOA algorithms to
planned obstacle, ultimately reaching the
solve this problem and compare the results to see
designated goal.
which one works best.

Fig. 5 Optimizing robot paths: PSO algorithm's function

Fig. 3 Path planning problem used in this research
evaluation graph
A. Analysing Optimization Results: Robot Path In Figure 5, you can see that the cost value
Planning with PSO Algorithm dropped quickly when we used the PSO algorithm
Here we will use a particle swarm optimization
algorithm for robot path planning. First of all, to
109
with specific settings from Table 1. The best
recorded PSO algorithm cost value is 7.5558.
B. Analysing Optimization Results: Robot Path
Planning with GOA Algorithm
For robot path planning, we will employ the
Grasshopper Optimization Algorithm (GOA). To
utilize the GOA algorithm, we need to establish the
values of specific hyperparameters beforehand, as
outlined in Table 2.
Table 2. The GOA algorithm's hyperparameter values
cMax cMin MaxIt nPop
1 0.01 1000 500
We applied the GOA algorithm with the
Fig. 7 Optimizing robot paths: GOA algorithm's function
parameters outlined in Table 2 to guide the robot evaluation graph
around obstacles and achieve the goal illustrated in
Figure 3. The ensuing results are detailed below. Figure 5 highlights the rapid reduction in cost
when employing the GOA algorithm with specified
settings from Table 2. The best-recorded cost value
for the GOA algorithm is 7.5421.
V. DISCUSSION
This section delves into discussion details
regarding robot path planning, focusing on the
application of the Grasshopper Optimization
Algorithm, in the preceding sections. A
comparative analysis is conducted, contrasting our
approach's outcomes with those of other
researchers who tackled path-planning challenges
utilizing the obstacles depicted in Figure 3.
In this article [30], the authors explored robot
Fig. 6 The best solution obtained by the grasshopper path planning employing the Gray Wolf
optimization algorithm Optimization (GWO) algorithm with the same
obstacles. Their study involved running the GWO
Figure 6 shows that the GOA algorithm smoothly
algorithm with 50 populations and 1000 iterations,
went around the obstacle as planned and reached
revealing the best cost to be 7.652. In our
the goal.
implementation, utilizing the Grasshopper
Optimization Algorithm, we achieved a superior
result with a best recorded cost of 7.5421. This
disparity underscores the effectiveness of the
grasshopper optimization algorithm applied in our
study.
VI. CONCLUSION AND FUTURE WORKS
Robot path planning is one of the most difficult
problems in the field of optimization. To solve this
problem, you can use the classical methods that
exist in this field, but most of these methods are
not very applicable due to the complexity of the
problem. Therefore, meta-heuristic methods can be

110
used to solve such problems and find an optimal Instrumentation & Information Systems (EIIS), pp.
solution for them. 1-4, 2017. pp. 1-4, 2017.
[9] Lyu D., Chen Z., Cai Z., and Piao S., "Robot path
In this article, we took on the challenge of robot planning by leveraging the graph-encoded Floyd
path planning using two different methods: Particle algorithm," Future Generation Computer Systems,
Swarm Optimization (PSO) and the Grasshopper vol. 122, no. pp. 204-208, 2021.
Optimization Algorithm (GOA). Both were trying [10] Li H., Zhao T., and Dian S., "Forward search
to find the shortest path. Our results showed that optimization and subgoal-based hybrid path
planning to shorten and smooth global path for
with PSO, we got a best cost value of 7.5558. On mobile robots," Knowledge-Based Systems, vol. no.
the other hand, GOA did a bit better with a best pp. 110034, 2022.
cost value of 7.5421. Looking at these results, it's [11] Bae H., Kim G., Kim J., Qian D., and Lee S., "Multi-
clear that the GOA method is better at handling the robot path planning method using reinforcement
tricky parts of robot path planning. learning," Applied sciences, vol. 9, no. 15, pp. 3057,
2019.
In this article, our main aim is to get the best [12] Ni J., Zhang Z., Su B., Fan X., and Liang W., "A bio-
results for planning robot paths using the inspired neural network based PSO method for robot
grasshopper optimization algorithm. In the future, path planning," 2017 13th International Conference
researchers can explore solving the robot path on Natural Computation, Fuzzy Systems and
planning puzzle using other smart algorithms like Knowledge Discovery (ICNC-FSKD), pp. 234-238,
2017. pp. 234-238, 2017.
MOPSO, MOGOA, SPEA-II, and MODE. This [13] Shareef A. and Al-Darraji S., "Grasshopper optimization
way, we can see how the results compare and algorithm based path planning for autonomous
maybe make the model even better. Also, trying mobile robot," Bulletin of Electrical Engineering
out our method on different sets of data will help and Informatics, vol. 11, no. 6, pp. 3551-3561,
us see how well it works for various situations. 2022.
[14] Qin Y.-Q., Sun D.-B., Li N., and Cen Y.-G., "Path
planning for mobile robot using the particle swarm
optimization with mutation operator," Proceedings
REFERENCES
of 2004 international conference on machine
[1] Koubaa A., Bennaceur H., Chaari I., Trigui S., Ammar learning and cybernetics (IEEE Cat. No. 04EX826),
A., Sriti M.-F., Alajlan M., Cheikhrouhou O., Javed pp. 2473-2478, 2004. pp. 2473-2478, 2004.
Y., and Koubaa A., "Introduction to mobile robot [15] Siegwart R., Nourbakhsh I.R., and Scaramuzza D.,
path planning," Robot path planning and Introduction to autonomous mobile robots, MIT
cooperation: foundations, algorithms and press, 2011.
experimentations, vol. no. pp. 3-12, 2018. [16] Klancar G., Zdesar A., Blazic S., and Skrjanc I.,
[2] Alexopoulos C. and Griffin P.M., "Path planning for a Wheeled mobile robotics: from fundamentals
mobile robot," IEEE Transactions on systems, man, towards autonomous systems, Butterworth-
and cybernetics, vol. 22, no. 2, pp. 318-322, 1992. Heinemann, 2017.
[3] Sariff N. and Buniyamin N., "An overview of [17] Koubaa A., Bennaceur H., Chaari I., Trigui S., Ammar
autonomous mobile robot path planning algorithms," A., Sriti M.-F., Alajlan M., Cheikhrouhou O., and
2006 4th student conference on research and Javed Y., Introduction to mobile robot path
development, pp. 183-188, 2006. pp. 183-188, 2006. planning, Springer, 2018.
[4] Panda M., Das B., Subudhi B., and Pati B.B., "A [18] Xue Y. and Sun J.-Q., "Solving the path planning
comprehensive review of path planning algorithms problem in mobile robotics with the multi-objective
for autonomous underwater vehicles," International evolutionary algorithm," Applied sciences, vol. 8,
Journal of Automation and Computing, vol. 17, no. no. 9, pp. 1425, 2018.
3, pp. 321-352, 2020. [19] Hussain K., Mohd Salleh M.N., Cheng S., and Shi Y.,
[5] Yang X.-S., Nature-inspired metaheuristic algorithms, "Metaheuristic research: a comprehensive survey,"
Luniver press, 2010. Artificial intelligence review, vol. 52, no. pp. 2191-
[6] Kennedy J. and Eberhart R., "Particle swarm 2233, 2019.
optimization," Proceedings of ICNN'95- [20] Burke E.K., Burke E.K., Kendall G., and Kendall G.,
international conference on neural networks, pp. Search methodologies: introductory tutorials in
1942-1948, 1995. pp. 1942-1948, 1995. optimization and decision support techniques,
[7] Saremi S., Mirjalili S., and Lewis A., "Grasshopper Springer, 2014.
optimisation algorithm: theory and application," [21] Yousefi T. and Aktaş Ö., "New hybrid segmentation
Advances in Engineering Software, vol. 105, no. pp. algorithm: UNet-GOA," PeerJ Computer Science,
30-47, 2017. vol. 9, no. pp. e1499, 2023.
[8] Yongxiang Z. and Lei Z., "Improvement and application [22] Sengupta R.N., Gupta A., and Dutta J., Decision
of heuristic search in multi-robot path planning," sciences: theory and practice, Crc Press, 2016.
2017 First International Conference on Electronics

111
[23] Koziel S. and Yang X.-S., Computational optimization,
methods and algorithms, Springer, 2011.
[24] Li C., Yang S., and Nguyen T.T., "A self-learning
particle swarm optimizer for global optimization
problems," IEEE Transactions on Systems, Man,
and Cybernetics, Part B (Cybernetics), vol. 42, no.
3, pp. 627-646, 2011.
[25] Wang D., Tan D., and Liu L., "Particle swarm
optimization algorithm: an overview," Soft
computing, vol. 22, no. 2, pp. 387-408, 2018.
[26] Tran B., Xue B., and Zhang M., "Overview of particle
swarm optimisation for feature selection in
classification," Asia-Pacific conference on simulated
evolution and learning, pp. 605-617, 2014. pp. 605-
617, 2014.
[27] Shi Y. and Eberhart R., "A modified particle swarm
optimizer," 1998 IEEE international conference on
evolutionary computation proceedings. IEEE world
congress on computational intelligence (Cat. No.
98TH8360), pp. 69-73, 1998. pp. 69-73, 1998.
[28] Ewees A.A., Abd Elaziz M., Alameer Z., Ye H., and
Jianhua Z., "Improving multilayer perceptron neural
network using chaotic grasshopper optimization
algorithm to forecast iron ore price volatility,"
Resources Policy, vol. 65, no. pp. 101555, 2020.
[29] Beni G. and Wang J., Swarm intelligence in cellular
robotic systems, Springer, Berlin, 1993.
[30] Doğan L. and Yüzgeç U., "Robot path planning using
gray wolf optimizer," Proceedings-International
Conference on Advanced Technologies, Computer
Engineering and Science (ICATCES’18), pp. 2018.
pp. 2018.

112

View publication stats