Login (http://www.tjprc.org/login.php) | Register (http://www.tjprc.org/registration.

php) Seleccionar idioma ▼

 (http://www.linkedin.com/groups/Transstellar-Journal-Publications-Research-Consultancy-3975427) 
(http://twitter.com/#!/TJPRCPvtLmt)  (http://www.facebook.com/TJPRC) (http://tjprc.org/rss.xml) 
(https://plus.google.com/101820340392906499167)

|  (http://www.tjprc.org/index.php)  (http://www.tjprc.org/contact-us.php)  (http://www.tjprc.org/sitemap.php) 

(http://www.tjprc.org/index.php)
Call For Papers
(http://www.tjprc.org/papers/journal-impact- Ongoing Issue(Non-Scopus Journal): May - Jun 2020
factor) Submission Deadline: 08 May 2020
Crossref Indexed(DOI), Scopus Indexed(IJMPERD) with ISSN & Submission email id: [email protected]
Impact Factor(JCC)
(http://www.tjprc.org/ugc-
T
approved-journals)
(http://www.tjprc.org/papers/journal-impact-
factor)
Scopus Indexed Journal Details (https://www.scopus.com/sourceid/21100814505)

International Journal of Mechanical and Production Engineering Research and Development (IJMPERD)

ISSN(Print) : 2249-6890 ISSN(Online) : 2249-8001

Impact Factor(JCC) : 8.8746 NAAS Rating : 3.11

IBI Factor : 3.2

Keyword(by author, title)

Journal Type

International Journal
 Journal name

International Journal of Mechanical and Production Engineering Research and Development

Published Date From

to to

Search Reset (view-archives.php)

Quick Menus

 About (http://www.tjprc.org/journals/journal-of-mechanical-engineering)

 Submit Manuscripts (http://www.tjprc.org/submit-papers.php?year=2020&jtype=2&id=67)

 Author Guidelines (http://www.tjprc.org/page-detail/author-guidelines37/2020/2/67)

 Editorial Board (http://www.tjprc.org/journal-detail/journal-of-mechanical-engineering/2/67/editors)

 Archive (http://www.tjprc.org/view-archives.php?year=2020&journal=67&id=67&jtype=2&details=archives)

 Copyright Transfer Agreement (http://www.tjprc.org/trans_global.pdf)

 Publication Ethics (http://www.tjprc.org/publication_ethics.php?jtype=2&id=67)

 

Vol
Sl. Issue Page
- Paper Id Title Abstract Author
No. Date No
Issue

216 Jun-30 10– IJMPERDJU A Numerical Study on Thermal Stress Abstract Dr. Ch. Srinivasa 305–
1 2020 3 N202029 Analysis of A Micro Electro Mechanical (http://ww Rao 314
Systems (Mems) Bimetallic Actuator w.tjprc.org
(view_paper.php?id=12680) /publishpa
DOI : 10.24247/ijmperdjun202029 pers/2-67-
158859228
8-
abs29IJMP
ERDJUN20
2029.pdf)

216 Jun-30 10– IJMPERDJU Experimentnal Approach to the Design Abstract Ricardo Castillo et 315–
2 2020 3 N202030 of the Links of a Delta Robot (http://ww al., 328
(view_paper.php?id=12682) w.tjprc.org
DOI : 10.24247/ijmperdjun202030 /publishpa
pers/2-67-
158859352
3-
abs.30.IJM
PERDJUN2
02030.pdf)
 Vol
Sl. Issue Page
- Paper Id Title Abstract Author
No. Date No
Issue

216 Jun-30 10– IJMPERDJU Vibratonal Analysis of Rotor Dynamic Abstract P. Ravinder Reddy 329–
3 2020 3 N202031 System using FEA (view_paper.php? (http://ww et al., 340
id=12683) w.tjprc.org
/publishpa
DOI : 10.24247/ijmperdjun202031 pers/2-67-
158859664
0-
abs.31.IJM
PERDJUN2
02031.pdf)

<< (view-archives.php?keyword=&from_date=&to_date=&id=67&jtype=2&journal=67&page=1)

< (view-archives.php?keyword=&from_date=&to_date=&id=67&jtype=2&journal=67&page=144)

Page 145 of 145 Go > >>


(http://www.tjprc.org/papers/peer-
reviewed-journals)

PUBLISHING & PRINTING ACTIVITIES

(http://www.tjprc.org/papers/peer-reviewed-journals)


(http://www.tjprc.org/papers/research-
board21)

RESEARCH BOARD

(http://www.tjprc.org/papers/research-board21)


(http://www.tjprc.org/papers/why-
publish-and-

WHY PUBLISH research-with-us)

AND RESEARCH WITH US?
(http://www.tjprc.org/papers/why-publish-and-research-with-us)


(http://www.tjprc.org/papers/journal-
of-knowledge-
management)
EDITORIAL PROCESS & PUBLICATION

(http://www.tjprc.org/papers/journal-of-knowledge-management)


(http://www.tjprc.org/papers/paper-
publishing-sites)

PUBLISHING ETHICS

(http://www.tjprc.org/papers/paper-publishing-sites)


(http://www.tjprc.org/services/proofreading-
services)

PROOFREADING SERVICES

(http://www.tjprc.org/services/proofreading-services)

003398716 ()

0.07 2017
CiteScore

(https://www.scimagojr.com/journalsearch.php?
q=21100814505&tip=sid&clean=0)

11th percentile
Powered by
International Journal of Mechanical and Production
Engineering Research and Development (IJMPERD)
ISSN (P): 2249–6890; ISSN (E): 2249–8001
Vol. 10, Issue 3, Jun 2020, 315–328
© TJPRC Pvt. Ltd.

EXPERIMENTAL
AL APPROACH TO THE DESIGN OF THE LINKS OF A DELTA
ROBOT

RICARDO CASTILLO,
CASTILLO OSCAR F. AVILES & MAURICIO F. MAULEDOUX
Research Scholar, DAVINCI Research Group, Department of Mechatronics Engineering, Universidad Militar
Nueva Granada, Bogota, Colombia
ABSTRACT

This paper shows the mathematical development of a Delta parallel robot of four Degrees Of Freedom (DOF) to obtain
the length of its links, solving its direct and inverse kinematics and the Jacobian matrix. Also, an algorithm to find the
workspace’s robot is posed and finally, the dimension links are obtained by six different kinds of multi-objective
multi genetic
algorithm to maximize the workspace and dexterity of the robot.
robot

KEYWORDS: Delta Parallel Robot, Direct and Inverse Kinematic, Jacobian Matrix, Multi-Objective
Multi Objective Genetic Algorithm,
Workspace & Dexterity

Original Article
Received: Feb 25, 2020; Accepted: Mar 17, 2019; Published: May 04, 2020; Paper Id.: IJMPERDJUN202030
IJMPERD

1. INTRODUCTION

A robot is designed to be controlled by a computer or similar device to be able to perform multiple tasks by simply
changing its programming [1]. In the beginning, the robots performed simple and monotonous tasks in work
environments dangerous to man, achieving a reduction in the costs and manufacturing times of products and
increasing their quality and productivity [2]. Currently, robots are essential in industrial processes, being useful in
operations such as welding, painting, machining operations (drilling, milling, grooving, etc.), grinding, polishing,
brushing, cutting, product quality control, among others [3].

Until a few years ago, open-chain

chain manipulators were
were responsible for performing the tasks named above. A
serial manipulator is formed from a single kinematic chain and its mechanical structure resembles that of the human
arm (Figure 1). In these robots, the actuators are located along with its mechanical structure which generates an
increase in the inertia that the robot must overcome to move, that is, the more actuators the robot has, the greater the
load on it, affecting the stability of the whole mechanism. In addition to this, each actuator increases the
t probability
of generating a position and orientation error of the final effector [4].

Figure 1: Serial Manipulator.

www.tjprc.org SCOPUS Indexed Journal [email protected]

316 Ricardo Castillo, Oscar F. Aviles
Avil & Mauricio F. Mauledoux

Parallel manipulators emerged (Figure 2), where two or more kinematic chains support the mobile platform
increasing rigidity, stability, load capacity, the accuracy of position and orientation of the final effector, speed and the
acceleration of robot movements [5]. These advantages make these robots suitable for applications such as:

• Pick and place for packaging, packaging or quality control operations in the food industry.

• Micro movement robots.

• Motion simulators.

• Haptic interfaces.

Figure 2: Parallel Manipulator.

However, compared to serial manipulators, parallel manipulators have three disadvantages: difficulty in the
development of their kinematics, less workspace, and increased singularities, which hinders the performance of the robot in
its environment [5].. Over the years, different solutions have been
been proposed to the development of the kinematics of
parallel robots such as the Delta type (Figure 2). One way to solve the kinematics is through an analysis of its geometry
[6], [7] or using software such as ADAMS [8].

To solve the problem of reduced workspace,

workspace, new forms of parallel robots have been developed, where revolution
joints are combined with binoculars [9] or the prismatic joints replace those of revolution [10], [11]. These new
configurations achieve a greater workspace, but the robot's movement speed is reduced because the linear actuators cannot
reach the speeds that the revolution actuators reach. Algorithms have also been implemented to maximize the workload of
parallel robots using genetic algorithms [5], [9], [12], [13], the SQP method [14], the CRS method [15], between others.

The objective of this work is to show how the workspace of a 4 DOF Delta robot can be optimized using multi-
multi
objective genetic algorithms. Section 2 presents the solution of direct kinematics, inverse kinematics and the
th Jacobian
matrix of the robot. Section 3 presents an algorithm to find the robot's workspace. In section 4,
4 you will find the
dimensions of the robot links to optimize your workspace. Finally, section 5 presents the conclusions.

1. DELTA ROBOT KINEMATICS

To obtain a mathematical model that describes the kinematics of the robot, its geometry must first be analyzed. As seen in
Figure 3, the Delta robot consists of a fixed base, three servomotors , y and at the same distance from the center

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the
he Design of the Links of a Delta Robot 317

of the base and separated120º each other, a mobile base and three kinematic chains, each consisting of two links, which
are connected employing universal joints (U). The reference system is located in the center of the fixed base, with the
axis perpendicularr to the axis of the servomotor and the axis is perpendicular to the base.

Figure 3: Delta Robot Structure.

In Figure 4, it can be seen that the link is attached to the fixed base through the joint and to the link by
the joint . The link joins the mobile platform through articulation . Finally, is the angle between link and .
The center of the mobile base represents the position of the Delta robot ! " # $% .
&

The distance between and , and is and the distance between and , or is )

Figure 4: Structure of the Delta Robot.

Below are the vectors that will be used to calculate the kinematics of the robot1:

(1)

Inverse Kinematics

Inverse kinematics allows knowing the variables of the actuators if the position of the mobile base is known. Thanks to the
symmetry of the robot, each kinematic chain can be analyzed independently.
independently As can be seen in Figure 5 there are in total
there are nine joint angles: , , for ' 1,2,3 associated to the manipulator, being , , the variables of the
actuator joints.

sen s , and cos c

www.tjprc.org SCOPUS Indexed Journal [email protected]

318 Ricardo Castillo, Oscar F. Aviles
Avil & Mauricio F. Mauledoux

1
Notation. sen s , and cos
cos c

Figure 5: Description of the Angles:

Angles: a) Front View and) Lateral View.

It is possible to form a triangle with vertices as shown in Figure 6 and the magnitude can be found by
means of the cosine theorem:

* +2 cos, (7)

* +2 cos,

Figure 6: Vector Representation of the First Kinematic Chain.

Chain

Applying the definition of the scalar product is found c, :

0000000001 000000001
c,
- . /. ∙- . 3.
45 46
(8)

000000001 00000001
7 7∙ 7 7
c,
Replacing Eq. (8) in Eq. (7) you get:

+ *2 c 8 c9 * : s9 *)+ *2 s ; (9)

The magnitude of the vector Eq. (6) is:

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the Design of the Links of a Delta Robot 319

< 8 * c9 ) + = *< : * s9 ) + = * ; (10)

Equating Eq. (9) and Eq. (10) you get an equation of the form:

>s *?c @ (11)

Where,

@> 2 ;

@? 2 8 @9 * : A9 *)+

2 2
@@ B * @9' ) + C *B * A9' ) + C * 2
* 2
1 + 2
2

It is possible to rewrite Eq. (11) taking into account the trigonometric identity of sum of sines and cosines:

senD E L + tanD B C
FFG FKG

HFIG6 JFKG6 FIG

(12)

In this way@@the values of D , @? ,

senD N P + tan Q R
@>
if the position of the mobile base is known.
O@> * @?
Direct Kinematics

The problem of direct kinematics consists in determining the position and orientation of the final effector of the robot from
a reference coordinate system. The magnitude of the vector Eq. (5) is :

< 8 + c9 * c +) = *< : + s9 * c +) = * $ + s (13)

Eq. (13) represents three nonlinear equations that, when solved, allow us to find the position of the mobile base if
the angles are known , , .

/J-ST
: 3
(14)

Where, * ;
:
> +? > +@
+
2> + ? 2> + @

2 @ +> 2 ? +>
+
2> + @ 2> + ?

√3? √3@
*
2> + ? 2> + @

8 V+W ;

And

> +? √3?
V +
2> + ? 2> + @

2 ? +> √3?
W *
2> + ? 2> + ?

www.tjprc.org SCOPUS Indexed Journal [email protected]

320 Ricardo Castillo, Oscar F. Aviles & Mauricio F. Mauledoux

DKJOK 6 DXIF
; I
(16)

and > + ? √3? 2 ? +> √3? +? * √? + 4>@

V + W *
2> + ? 2> + @ 2> + ? 2> + ? ;
2>
> W *Q R *1

? 2Q + VW + > W + > R

@ V * Q R * 2> V + >

Additionally,

> )+ + c > s > +> +>

? +)* c ? s ? +? +?

@ +)* c @ s @ +@ +@

Jacobian Matrix

A closed loop equation of the 'Z[ kinematic chain can be: 000001 * 00000017 000000017 * 000000001 00000001
7 7 * 7 7 . If you derive this expression

with respect to time and the result is expressed in the system is obtained:

0001
\] 00000017 × 000000001
^ 7 7*^ 00000017 × 00000001
7 7 (17)

00000017 represents the velocities of the passive joints, therefore, they are eliminated by multiplying both
In Eq. (17) ^
sides of the equation by 00000001
7 7 what results:

\] ∙ 00000001
0001 7 7
00000001
7 7 <^00000017 × 000000001
7 7= (18)

where:

c c *
00000001
7 7 ` s a
c s *

\]8 c9 * \]: s9
0001
\] `\]: c9 + \]8 s9 a
\];

0
^
00000017 b+ c d
0

c
00000001
7 b 0 d
s

Solving Eq. (18) and simplifying you get:

e" \]8 * e" \]: * e" \]; ef c (19)

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the Design of the Links of a Delta Robot 321

where:

e" c c * c9 + s s9

e" c c * s9 + s c9

e" c s *

ef s c
The expression obtained in Eq. (19) represents a system of three equations that can be written in the form:

e" e" e" \]8 ef 0 0 c

be" e" e" d `\]: a b0 ef c
0 d` a
e" e" e" \]; 0 0 ef c
(20)

So, the Jacobian matrix of the Delta robot is:

ef 0 0 D e" e" e"

g gf D
g" b0 ef 0 d be" e" e" d
0 0 ef e" e" e"
(21)

It is said that there is a singularity of inverse kinematics when the determinant gf is equal to zero; it is said that
there is a singularity of inverse kinematics when the determinant of g" is equal to zero; the singularities are said to be
combined when the determinants of gf and g" are zero.

Robot Dexterity

The Jacobian matrix allows identifying unwanted configurations in a Delta robot (singularities). Additionally, it allows
knowing the skill level of each robot configuration through the condition number of the Jacobian matrix:

λ ‖J‖ ∙ ‖J D ‖ (22)

The local skill of a parallel robot is defined as the inverse of the condition number of the Jacobian matrix [16]:

ς
l
(23)

1 If the value of m is zero it is because the robot is in a unique configuration and if the value of m is one, the robot
ς
λ
configuration is known as isotropic [17]. Given that the m index allows analyzing only each configuration of the parallel
robot, Gosselin, and Angeles in 1991, they proposed an index that allows measuring the dexterity of a robot in its entire
workspace (global index):

os pqr
n
os qr
(24)

ov mtu
Where the numerator ov mtu represents the sum of the local skills of the Delta robot throughout its workspace
n
and the tu
ovdenominator ov tu is the value of the robot's workspace.

2. DELTA ROBOT WORKSPACE

The workspace of a robot is considered as the region in the three-dimensional Cartesian space that can be reached by a
point of its final effector [12], in the case of a Delta robot, it is the region in the three-dimensional space that can reach one

www.tjprc.org SCOPUS Indexed Journal [email protected]

322 Ricardo Castillo, Oscar F. Aviles & Mauricio F. Mauledoux

point of your mobile base. Designing a Delta robot for maximum workspace does not guarantee that the robot will be
optimal for practical applications and it is feasible that it becomes a manipulator with unwanted kinematic characteristics
such as low dexterity and maneuverability [18]. For this reason, in this article, an algorithm was used to calculate the exact
amount of points in the space that is capable of reaching the robot's mobile base. The algorithm is based on the solution of
direct kinematics for all possible combinations of the actuators, obtaining in each case a point in the space, which is
evaluated to know whether it belongs to the robot's workspace. The algorithm also allows you to find the overall skill of
the robot's workspace.

First, the local dexterity index of each point that belongs to your workspace is calculated, then all the local
indexes are added and divided by the number of points that the final effector of the robot can reach in the space.

3. OPTIMIZATION OF THE WORKSPACE

It seeks to find a set of decision vectors; whose elements are the dimensions of the links and that maximize the workspace
and skill of the Delta robot. In this way, the objective functions are the number of points that make up the robot's
workspace and the global skill index raised by Gosselin and Angeles.

The dimensions of the links must meet the following restrictions:

<
100 < < 350
360 < < 600
* < 900

Figure 7 shows the problem-solving space where the blue dots represent the possible dimensions of the links and
the red dots are dimensions that do not meet the restrictions outlined above. To find the set consisting of the best solutions,
six different methods based on genetic algorithms were proposed:

Figure 7: Problem Solving Space.

To find the set consisting of the best solutions, six different methods based on genetic algorithms were proposed:

• RWGA: Random Weight Genetic Algorithm.

• FWGA: Fixed Weight Genetic Algorithm.
• AWGA: Adaptive Weight Genetic Algorithm.
• NSGA: Non-dominated Sorting Genetic Algorithm.

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the Design of the Links of a Delta Robot 323

• NPGA: Niched Pareto Genetic Algorithm.

• MOFGA: Multiplied Objectives Genetic Algorithm.

The first five algorithms are known in the view that many investigations have already been developed around
them. The last algorithm is a method proposed in this article where the adaptability value of each individual is found by
multiplying the two values of the objective functions to preserve the individuals with the highest amount of points they can
reach and the overall skill index higher. Thanks to the fact that the six algorithms are based on genetic algorithms, the
following nine steps can be followed to find the solution set of the problem:

Step 1: Generate an initial population of 50 individuals, verifying that the individuals comply with the restrictions
of the links y .

Step 2: Evaluate each individual in the two objective functions, that is, obtain the overall skill index and the
number of points that make up the Delta robot's workspace for each solution.

Step 3: A tentative set of optimal Pareto solutions is formed, where half corresponds to the ten individuals of the
current population with the most points in the space that can reach the mobile base and the other half are the ten individuals
with the best overall skill index.

Step 4: An adaptive value is assigned to each individual.

Step 5: Through the selection operator, some individuals are chosen as parents. The greater the adaptability of an
individual, the greater the probability of being chosen as a parent. To have more solution options, three selection methods
(truncation, stochastic universal sampling, and roulette) were used for each algorithm.

Step 6: { new individuals (offspring) are generated through the intermediate crossing operator.

Step 7: Through the mutation operator, some individuals are chosen randomly to generate small changes.

Step 8: Some individuals are randomly selected from the population to be changed by individuals from the Pareto
tentative set.

Step 9: Steps 2 through 9 are repeated until a total of 100 generations are completed.

Once the 18 algorithms were executed, six multi-objective genetic algorithms each for three selection methods,
the solution sets were grouped according to the selection method, see Figure 8, Figure 9 and Figure 10.

Figure 8: Set of Solutions with the Truncation Method.

www.tjprc.org SCOPUS Indexed Journal [email protected]

324 Ricardo Castillo, Oscar F. Aviles & Mauricio F. Mauledoux

Figure 9: Set of Solutions with the SUS Method.

Figure 10: Set of Solutions with the Roulette Method.

To find the final set of solutions, these three sets were united into one and its optimal Pareto boundary (non-
dominated solutions) was identified, as shown in Figure 11.

Of that Pareto optimal border, individuals with the highest overall skill (greater than 0.47) and the highest number
of points in the workspace (more than 200,000 points) are those of interest to obtain the dimensions of the links of the
Delta robot Analyzing each of the individuals and their new values of objective functions, the dimensions of the robot links
are: 260mm y 420mm.

Figure 11: Pareto Front of the Final Set of Solutions,

MATLAB.

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the Design of the Links of a Delta Robot 325

Figures 12 to 15 show the different views of the real workspace of the Delta robot, with the dimensions of the
links found. The data of this workspace are:

• Points in space: 259656

• Optimum points: 146273
• Global skill: 0,556
• Minimum reach: +342,8015 mm, +296,9611 mm y 105,7695 mm
• Maximum reach: 342,4004 ••, 296,9611 •• y 677,0132 ••

Figure 12: Workspace of the Delta Robot. Figure 13: Top View of the Robot Workspace.

Figure 14: View of the Workspace of the Robot in XZ Figure 15: View of the Workspace of the Robot in
plane YZ

5. CONCLUSIONS

The direct and inverse kinematics of a Delta robot can be solved by analyzing the geometry of the robot and posing
equations that relate its dimensions, the angular position of the actuators and the position of the mobile base. Some of the
most important characteristics to consider when designing a parallel robot are its workspace and its skill, given that these
are the two shortcomings of these robots compared to serial manipulators.

As the workspace of parallel robots is composed of irregular volumes, it is necessary to rely on software, Matlab
in this case, to be able to visualize all the points that the robot's mobile platform is able to reach. It was once again verified
that genetic algorithms are very useful when solving optimization problems thanks to their robustness and easy
implementation. In this case, not a single algorithm was implemented, but 18 algorithms to find the dimensions of the robot
links that best show a balance between workspace and skill. The results obtained were dimensions of 260 mm and 420 mm

www.tjprc.org SCOPUS Indexed Journal [email protected]

326 Ricardo Castillo, Oscar F. Aviles & Mauricio F. Mauledoux

with which the robot can reach almost 260 000 points in the space of which 146 273 points are optimal, that is, they
comply with a local skill factor greater than 0.5. Also, the overall skill of the robot is 0.556, which meets the objective of
designing a robot to maximize its workspace and its skill.

ACKNOWLEDGEMENT

Authors gratefully acknowledge the support of Lucas Urrea Mantilla, Sergio Alejandro Medina for their valuable support.
Also, the research for paper was supported by Universidad Militar Nueva Granada, through the project ING-IMP-2657
(“Platform for the evaluation of therapeutic adherence and its influence on the efficiency of stabilometry therapies”).

REFERENCES

1. S. B. Niku, Introduction to Robotics: Analysis, Control, Applications, 2a ed. Wiley, 2010.

2. L.-W. Tsai, Robot Analysis: The Mechanics of Serial and Parallel Manipulators, 1a. ed. Wiley-Interscience, 1999.

3. M. Groover, M. Weiss, R. Nagel, y N. Odrey, Robótica industrial: tecnología, programación y aplicaciones. Madrid, España:
McGrawHill, 1990.

4. D. Zhang, Parallel Robotic Machine Tools. Springer, 2009.

5. Y. Lou, G. Liu, J. Xu, y Z. Li, «A general approach for optimal kinematic design of parallel manipulators», presented at the
2004 IEEE International Conference on Robotics and Automation (ICRA’04), New Orleans, LA, 2004, vol. 4, pp. 3659– 3664.

6. L.-W. Tsai, G. C. Walsh, y R. E. Stamper, «Kinematics of a novel three DOF translational platform», presented at the 1996
IEEE International Conference on Robotics and Automation, Minneapolis, Minnesota, 1996, vol. 4, pp. 3446–3451.

7. J. Hong y M. Yamamoto, «A calculation method of the reaction force and moment for a Delta-type parallel link robot fixed
with a frame», Robotica, vol. 27, no. 4, pp. 579–587, jul. 2009.

8. Z. Jingjun, S. Lihong, G. Ruizhen, y L. Chaoyang, «The mathematical model and direct kinematics solution analysis of Delta
parallel robot», presented at the Second IEEE International Conference on Computer Science and Information Technology
(ICCSIT), Beijing, China, 2009, pp. 450–454.

9. S.-D. Stan, M. Manic, V. Maties, y R. Balan, «Kinematics analysis, design, and control of an Isoglide3 Parallel Robot
(IG3PR)», presented at the Thirtieth Annual Conference of IEEE Industrial Electronics (IECON 2008), Orlando, FL, 2008,
pp. 2636–2641.

10. Xin-Jun Liu y Jongwon Kim, «A new three-degree-of-freedom parallel manipulator», presented at the 2002 IEEE
International Conference on Robotics and Automation (ICRA’02), Washington, D.C., 2002, vol. 2, pp. 1155– 1160.

11. X.-J. Liu y J. Kim, «A new spatial three-DoF parallel manipulator with high rotational capability», IEEE/ASME Transactions
on Mechatronics, vol. 10, no. 5, pp. 502– 512, oct. 2005.

12. M. A. Laribi, L. Romdhane, y S. Zeghloul, «Analysis and dimensional synthesis of the DELTA robot for a prescribed
workspace», Mechanism and Machine Theory, vol. 42, no. 7, pp. 859–870, jul. 2007.

13. E. Martínez, C. Peña, y E. Yime, «Diseño óptimo de un robot paralelo con configuración Detla para aplicaciones educativas»,
Revista Educación en Ingeniería, no. 10, pp. 110–119, dic-2010.

14. M. Ceccarelli y C. Lanni, «A multi-objective optimum design of general 3R manipulators for prescribed workspace limits»,
Mechanism and Machine Theory, vol. 39, no. 2, pp. 119–132, feb. 2004.

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11
Experimental Approach to the
he Design of the Links of a Delta Robot 327

15. L. Yunjiang, L. Guanfeng, C. Ni, y L. Zexiang, «Optimal design of parallel manipulators for maximum effective regular
workspace», in 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2005), 2005, pp. 795–
795 800.

16. L. A. Silva, J. M. Sebastian, R. Saltaren, R. Aracil, y J. Sanpedro, «RoboTenis: optimal design of a parallel robot with high
performance», presented at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS ), 2005, pp.
2134– 2139.

17. M. Badescu, J. Morman, y C. Mavroidis, «Workspace optimization of 3-UPU

3 UPU parallel platforms with joint constraints», in
IEEE International Conference on Robotics and Automatio (ICRA ’02), Washington, D.C., 2002, vol. 4, pp. 3678: 3683.

18. R. E. Stamper, «A Three Degree of Freedom Parallel Manipulator with Only Translational Degrees of Freedom», University
of Maryland, College Park, Maryland, 1997

19. Ramprasad Kannemadugu & A. Lakshmi Devi, “Performance Improvement of Transmission System Using UPFC by GA and
PSO
O Algorithms”, International Journal of Electrical and Electronics Engineering (IJEEE), Vol. 3, Issue 1, pp. 1-8
1

20. Ranu Paul, “Stability Analysis of Critical Points to Control Growth of Tumor in an Immune-Tumor-Normal
Immune Cell-Drug
Model”, International Journal of Applied Mathematics & Statistical Sciences (IJAMSS), Vol. 5, Issue 6, pp. 43-52
43

21. K. Rajeev & K. Supraja, “A Study of Intelligent Controllers Application in Distributed Systems”, International Journal of
Mechanical Engineering (IJME), Vol. 5, Issue 4, pp.
pp 53-62

22. Amitvikram Nawalagatti & Prakash R. Kolhe, “A Comparative Study on Artificial Cognition and Advances in Artificial
Intelligence for Social-Human
Human Robot Interaction”, International Journal of Robotics Research and Development (IJRRD),
Vol. 8, Issue 1, pp. 1-10

AUTHOR’S PROFILE

Ricardo A. Castillo E. was born in Colombia in 1980. He received his B.Eng. in Mechatronics Engineering in
Universidad Militar Nueva Granada in 2004, and his M.Sc. and Ph.D. in Mechanical Engineering in the University of
Campinas: UNICAMP, Brazil in 2010 and 2015 respectively, working on coordination and indirect communication
strategies for mechatronic systems. Since 2005, he is a full-time
full time Professor and Researcher in the Department of
Mechatronics Engineering at Universidad Militar Nueva Granada (Bogotá:
(Bogotá Colombia). His current
curre research projects deal
with collaborative modular robotics, artificial intelligence and mobile autonomous robotics.

Oscar F Avilés S. He received the Engineer degree in Electronics and Specialist of Electronic Instrumentation of Antonio
Nariño University: UAN - in 1995 and 2002, respectively.. Master in Automatic Production Systems - Technological
University of Pereira - 2006 Colombia and PhD in Mechanical Engineering in Campinas State University:
University UNICAMP:

www.tjprc.org SCOPUS Indexed Journal [email protected]

328 Ricardo Castillo, Oscar F. Aviles
Avil & Mauricio F. Mauledoux

Brazil. Currently, he is Professor in the Department of Mechatronics Engineering in Universidad Militar Nueva Granada -
UMNG. He has experience in the areas of Electronic Engineering, with emphasis on electronic instrumentation and control
systems, acting mainly on the following topics: robotics, control and biomechatronics.

Mauricio Mauledoux He received the Engineer degree in Mechatronic engineering from the Universidad Militar Nueva
Granada, in 2005. In 2008 he was a student of the Master in Information
Information Technologies and Intelligent Systems in the St.
Petersburg State Polytechnic University, Russia, at the automatic and intelligent distributed control department. He was
then promoted to a PhD. In 2011, he received the PhD degree in Mathematical models, numerical methods and software
systems (Red Diploma) from the St. Petersburg State Polytechnic University, Russia. In 2012, he joined the Department of
Mechatronic Engineering, at Universidad Militar Nueva Granada, in Colombia, as an Assistant Professor. His
H current
research interests include Robotics, automatic control, Multi-agent
Multi agent Systems, Smart Grids, and Optimization.

Impact Factor (JCC): 8.8746 SCOPUS Indexed Journal NAAS Rating: 3.11