IAES International Journal of Robotics and Automation (IJRA)

Vol. 13, No. 2, June 2024, pp. 160~167

ISSN: 2722-2586, DOI: 10.11591/ijra.v13i2.pp160-167  160

Kinematic modelling of three link robot manipulator and joint

torque optimization using genetic algorithm in MATLAB

Neeta Sahay1, Subrata Chattopadhyay2

1
Department of Electrical and Electronics Engineering, Institute of Engineering and Management, Kolkata, India
2
Department of Electrical Engineering, National Institute of Technical Teachers’ Training and Research, Kolkata, India

Article Info ABSTRACT

Article history: This research article presents the non-linear dynamic of a three-link robotic
manipulator formulated by the Newton-Euler method. The planar
Received Jul 3, 2023 manipulator is composed of three links and three revolute joints rotating
Revised Mar 20, 2024 about the z-axis. The three nonlinear non-homogeneous dynamic equations
Accepted Apr 21, 2024 have been solved graphically with the help of MATLAB by phase variable
method. The work represents the graphical solution of the transient response
of angular position, and angular velocity of each link member for a
Keywords: predetermined interval of time. With the help of simulated value from
MATLAB, torque characteristics have been determined for different torque
Genetic algorithm ratios and optimum torque has been derived using a genetic algorithm to
MATLAB move the manipulator in a proper direction.
Newton-Euler
Optimization
Three-link manipulator This is an open access article under the CC BY-SA license.

Corresponding Author:
Neeta Sahay
Department of Electrical and Electronics Engineering, Institute of Engineering and Management
Kolkata, West Bengal, India
Email: [email protected]

1. INTRODUCTION
Robot manipulators can be assumed to be a chain of link mechanisms with highly nonlinear
dynamics. The forward and inverse kinematics are presented using the D-H convention and the deduced
mathematical model has been inferred by Newton-Euler approach. To obtain the optimal performance of the
robot manipulator, a precise dynamic model of the robot manipulator is required. Garg et al. [1] determined an
optimal path using a genetic algorithm and optimization was achieved using simulated annealing (SA).
Exhaustive simulation was conducted for different types of manipulators namely MELFA RV-1A having six
degrees of freedom and three planar revolute joints to find the distance between the end effector and the object
using an artificial neural network [2]. The payload of the mobile robotic arm is determined by a nonlinear
control law using optimal feedback [3] designed for a given trajectory task. This law is given by the solution
by the iterative method of a sequence of nonlinear Hamilton-Jacobi-Bellman equations. Aghanouri et al. [4]
developed a manipulator actuated by DC motors. The optimum path is derived by optimizing system
parameters to minimize performance indices including energy [5]. Talezadeh et al. [6] presented the nonlinear
dynamics of two-link manipulators for dynamic modeling of the system by the Lagrange equation of motion
where optimum control was adopted to analyse the motion of the manipulator. Some literature has presented
the whole-body dynamics of nonlinear equations of the hybrid cable-driven robots (HCDRs) where new
methods were developed to solve redundancy. Compared to the existing methods, torque optimization for
actuated and un-actuated joints can solve resolution problems as well as active satisfactory disturbance [7].
Baressi et al. [8] describe the kinematic modeling using the Denavit Hartenberg (DH) convention
and the dynamic modeling robotic arm has been presented by Lagrange-Euler (LE) and Newton-Euler

Journal homepage: http://ijra.iaescore.com

IAES Int J Rob & Autom ISSN: 2722-2586  161

algorithms. Different algorithms such as artificial intelligence, genetic algorithm, simulated annealing, and
differential evolution have been adopted and compared to provide the best results for minimization of torque.
Televnoy et al. [9] provides Lagrange equations of motion of a nonlinear matrix equation. The experiment
has been performed to find out the solution of a moving object between two points. The angular displacement,
speed, and acceleration for six links of the manipulator have been presented. Agustian et al. [10] present the
vision-based robotic manipulator where inverse kinematics using pseudo-inverse Jacobian (PIJ) and DH
forward kinematics have been adopted and controlled by a proportional derivative controller. The sorting task
depending on color was made to evaluate the error to implement the manipulator on a real system. The reviews
was given [11] on dynamic analysis and intelligent control techniques for flexible robot manipulators. A
comparative study of dynamic analysis and control strategies was presented for flexible manipulators.
Korayem et al. [12] have determined the non-linear dynamics and control of flexible mobile manipulators
focusing on the determination of maximum payload. Wu et al. [13] present the minimum actuator torque
range by torque optimization of a 3-DOF parallel manipulator. Two approaches, Lagrangian and primal-dual
neural network, were presented together to make real-time optimization of joint torque for kinematically
redundant manipulators [14], [15]. Gao et al. [16] proposed joint torque optimization of flexible manipulators
with redundant space and vibration suppression where the Lagrange method has been adopted to represent
the dynamics of the robot arm. Naghshineh and Keshmiri [17] investigated an over actuated system for
dynamic cost function by applying real time optimization on a cooperative robot. Wolniakowski et al. [18]
presented a method required for joint torque minimization of a serial manipulator to determine the optimal
task placement. Woolfrey et al. [19] described a control strategy for minimizing joint torque using null space
control of a redundant manipulator where the dynamic torque has been reduced by applying an external force
to the gripper element. Agbaraji et al. [20] presented the design of the manipulator by calculating and
analyzing the joint torques by evaluating performance in terms of speed and displacement of the arm based
on the predetermined values of the torques. Singh et al. [21] find the position vectors of the six-arm robot by
forward kinematics and joint angles by inverse kinematics in MATLAB with the help of a robotic toolbox.
Brandstotter et al. [22] and Petrenko et al. [23] presented a generalized closed-form solution of the dynamic
models of parallel robots using some simple Jacobian matrices where the dynamics of the legs were expressed
in the joint coordinates and that of the platform in the form of cartesian variables. Sun et al. [24] proposed the
methodology for deriving the closed-form inverse kinematic solutions of the 6-DOF robot on the position
level where the analytical inverse solution of all the joints was given out and compared with that of the forward
kinematic solution. The mathematical formulation and simulation of the two-link planar robot manipulator
with forward kinematics and dynamics were presented with the help of the D-H convention and Newton-Euler
method where the results in terms of joint angles and angular velocities have been presented graphically [25].
In this paper, the Newton-Euler formulation has been adopted to present the dynamics of the three-
link manipulator. The solution relating torque and angular displacement is obtained analytically by expressing
the three second-order non-linear non-homogeneous differential equations into six first-order differential
equations by phase variable method. The angular positions and speed of each link have been presented
graphically for constant torques of fixed duration using MATLAB. Then optimization is done using a genetic
algorithm to obtain minimum torque that can be applied to each link-joint actuator to obtain optimum results.

2. THREE LINK ROBOT MANIPULATOR

The length of the three links of the robotic manipulator is given by 𝑙1 , 𝑙2 𝑎𝑛𝑑 𝑙3 . Three joints are
named 𝐽1 , 𝐽2 , 𝐽3 , respectively, as in Figure 1. m1, m2, m3 are the mass of the first link, second link and third
link, respectively. Link parameters are considered to be 𝛼1 = 𝛼2 = 𝛼3 = 0. All the three joints are revolute
joints rotating about the z-axis, and all the links are rigid links. The angular rotations of each link-joint
combination are denoted as 𝜃1 ,𝜃2 , 𝜃3 .
The initial conditions are assumed to be

𝜔0 = 0 , 𝜔̇ 0 = 0 , 𝑣0 = 0

where 𝜔0 , initial angular speed, 𝜔̇ 0 , initial angular acceleration, 𝑣0 , initial linear velocity and 𝑣̇ 0 ,initial linear
acceleration. Therefore,

𝑣̇ 0 = [0 9.81 0]𝑇
𝑞𝑖 = [𝜃1 ,𝜃2 ,𝜃3 ];
𝑞̇ 𝑖 = [𝜃̇1 , 𝜃̇2 , 𝜃̇3 ]
𝑞̈ 𝑖 = [𝜃̈1 , 𝜃̈2 , 𝜃̈3 ];
Link variable: 𝐹𝑖 , 𝑓𝑖 , 𝑛𝑖 , 𝜏𝑖

Kinematic modelling of three link robot manipulator and joint torque optimization … (Neeta Sahay)
162  ISSN: 2722-2586

where 𝐹𝑖 is the total external force exerted at the center of i th link, 𝑓𝑖 is the force exerted on ith link by i-1st
link, 𝑛𝑖 is the moment acting on ith link by i-1st link, 𝜏𝑖 is the torque on ith joint.

Figure 1. Coordinate assignment of the robot manipulator

3. NEWTON-EULER FORMULATION FOR COMPUTATION OF JOINT TORQUE

Computational work has been carried out using Newton – Euler methodology for transformations of
three link coordinates to formulate the speeds of three-link manipulator. Initial speed of each link is assumed
to be zero. For the kinematic model, computation of each joint and link variables is required for computation
of each joint torque to move the manipulator in a desired direction. Forward kinematics and backward
kinematics of motion has been applied in the Newton – Euler approach as presented in (1) to (3). 𝜏1 ,𝜏2 ,𝜏3 are
the joint torque applied to the joint actuator for link i = 1, 2, 3, respectively.
1 2 1
𝜏3 = 𝑚3 𝑙1 𝑙3 [𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇1 ] + 𝑚3 𝑙2 𝑙3 [𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 +
2 2
1 1
𝑐𝑜𝑠 𝜃3 (𝜃̈1 + 𝜃̈2 )] + 𝑚3 𝑙3 2 (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑚3 𝑙3 𝑔 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) (1)
3 2

1 1
𝜏2 = 𝑚3 𝑙1 𝑙3 [𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇12 ] + 𝑚3 𝑙2 𝑙3 [𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 +
2 2
1 1
𝑐𝑜𝑠 𝜃3 (𝜃̈1 + 𝜃̈2 )] + 𝑚3 𝑙32 (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑚3 𝑙3 𝑔 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) + (𝑚3 +
3 2
1 1 1
𝑚 )𝑙 𝑙 [𝑐𝑜𝑠 𝜃2 𝜃̈1 + 𝑠𝑖𝑛 𝜃2 𝜃̇12 ] + (𝑚3 + 𝑚2 )𝑙22 (𝜃̈1 + 𝜃̈2 ) + 𝑚3 𝑙2 𝑙3 [𝑐𝑜𝑠 𝜃3 (𝜃̈1 +
2 2 1 2 3 2
1
𝜃̈2 + 𝜃̈3 ) − 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 ] + (𝑚3 + 𝑚2 )𝑙2 𝑔 𝑐𝑜𝑠( 𝜃1 + 𝜃2 ) (2)
2

1 2 1
𝜏1 = 𝑚3 𝑙1 𝑙3 [𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇1 ] + 𝑚3 𝑙2 𝑙3 [𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 +
2 2
1 1
𝑐𝑜𝑠 𝜃3 (𝜃̈1 + 𝜃̈2 )] + 𝑚3 𝑙32 (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑚3 𝑙3 𝑔 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) + (𝑚3 +
3 2
1 1 1
𝑚 )𝑙 𝑙 [𝑐𝑜𝑠 𝜃2 𝜃̈1 + 𝑠𝑖𝑛 𝜃2 𝜃̇12 ] + (𝑚3 + 𝑚2 )𝑙22 (𝜃̈1 + 𝜃̈2 ) + 𝑚3 𝑙2 𝑙3 [𝑐𝑜𝑠 𝜃3 (𝜃̈1 +
2 2 1 2 3 2
1
𝜃̈2 + 𝜃̈3 ) − 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 ] + (𝑚3 + 𝑚2 )𝑙2 𝑔 𝑐𝑜𝑠( 𝜃1 + 𝜃2 ) + (𝑚3 + 𝑚2 +
2
1 1
𝑚 )𝑙 2 𝜃̈ + (𝑚3 + 𝑚2 )𝑙1 𝑙2 [𝑐𝑜𝑠 𝜃2 (𝜃̈1 + 𝜃̈2 ) − 𝑠𝑖𝑛 𝜃2 (𝜃̇1 + 𝜃̇2 )2 ] −
3 1 1 1 2
1
𝑚3 𝑙1 𝑙3 [𝑠𝑖𝑛( 𝜃2 + 𝜃3 )(𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 − 𝑐𝑜𝑠( 𝜃2 + 𝜃3 )(𝜃̈1 + 𝜃̈2 + 𝜃̈3 )] + (𝑚3 +
2
1
𝑚2 + 𝑚1 )𝑙1 𝑔 𝑐𝑜𝑠 𝜃1 (3)
2

4. COMPUTED JOINT TORQUE CHARACTERISTICS OF EACH LINK-JOINT PAIR

The dynamic equations of motion of each link-joint coupling are highly non-linear in nature as shown
in (1) to (3). To solve the above equations of motion relating 𝜏𝑖 and 𝜃𝑖 (for ith link), phase variable model has
been adopted. The three non-linear non-homogeneous 2nd order differential equations are converted into six
first-order differential equations by phase variable method. Assuming 𝑧1 , 𝑧2 , 𝑧3 , 𝑧4 , 𝑧5 𝑎𝑛𝑑 𝑧6 are six phase
variables where:

𝑧1 = 𝜃1̇ (4)

𝑧2 = 𝜃2̇ (5)

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 160-167
IAES Int J Rob & Autom ISSN: 2722-2586  163

𝑧3 = 𝜃3̇ (6)

𝑧4 = 𝜃1 (7)

𝑧5 = 𝜃2 (8)

𝑧6 = 𝜃3 (9)

Therefore:

𝑧̇1 = 𝜃̈1 (10)

𝑧̇2 = 𝜃̈2 (11)

𝑧̇3 = 𝜃̈3 (12)

𝑧̇4 = 𝜃̇1 = 𝑧1 (13)

𝑧̇5 = 𝜃̇2 = 𝑧2 (14)

𝑧̇6 = 𝜃̇3 = 𝑧3 (15)

Non-dimensionalizing (1), (2) and (3), we have

1 1 2 1 1
𝜏3 = 𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇1 + 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 + 𝑐𝑜𝑠 𝜃3 (𝜃̈1 +
2 2 2 2
1 1
𝜃̈2 ) + (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) (16)
3 2

1 1 1 1
𝜏2 = 𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇12 + 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 + 𝑐𝑜𝑠 𝜃3 (𝜃̈1 +
2 2 2 2
1 1 3 3 4
𝜃̈2 ) + (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) + 𝑐𝑜𝑠 𝜃2 𝜃̈1 + 𝑠𝑖𝑛 𝜃2 𝜃̇12 + (𝜃̈1 +
3 2 2 2 3
1 1 3
𝜃̈2 ) + 𝑐𝑜𝑠 𝜃3 (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) − 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 + 𝑐𝑜𝑠( 𝜃1 + 𝜃2 ) (17)
2 2 2

1 1 2 1 1
𝜏1 = 𝑐𝑜𝑠( 𝜃2 + 𝜃3 )𝜃̈1 + 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )𝜃̇1 + 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 )2 + 𝑐𝑜𝑠 𝜃3 (𝜃̈1 +
2 2 2 2
1 1 3 3 4
𝜃̈2 ) + (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑐𝑜𝑠( 𝜃1 + 𝜃2 + 𝜃3 ) + 𝑐𝑜𝑠 𝜃2 𝜃̈1 + 𝑠𝑖𝑛 𝜃2 𝜃̇12 + (𝜃̈1 +
3 2 2 2 3
1 1 3 7
𝜃̈2 ) + 𝑐𝑜𝑠 𝜃3 (𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) − 𝑠𝑖𝑛 𝜃3 (𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 + 𝑐𝑜𝑠( 𝜃1 + 𝜃2 ) + 𝜃̈1 +
2 2 2 3
3 3 1
𝑐𝑜𝑠 𝜃2 (𝜃̈1 + 𝜃̈2 ) − 𝑠𝑖𝑛 𝜃2 (𝜃̇1 + 𝜃̇2 )2 − 𝑠𝑖𝑛( 𝜃2 + 𝜃3 )(𝜃̇1 + 𝜃̇2 + 𝜃̇3 )2 +
2 2 2
1 5
𝑐𝑜𝑠( 𝜃2 + 𝜃3 )(𝜃̈1 + 𝜃̈2 + 𝜃̈3 ) + 𝑐𝑜𝑠 𝜃1 (18)
2 2

Using (16) to (18), 𝜃̈1 , 𝜃̈2 and 𝜃̈3 can be expressed as a function of 𝜃1 , 𝜃2 , 𝜃3 and their derivatives. Therefore,
six first order equations can be formed from (10) to (15) and can be solved using MATLAB which gives the
relationship of angular displacement and applied torque for various torque ratios as shown in Figures 2 to 8.

Angular Rotation Vs. Applied Torque

Angular Rotation (rad)

4
3
2 theta1
1 theta2
0
-1 0 2 4 6 8 10 12
Applied Torque (Nm)

Figure 2. Angular position versus applied torque (τ1: τ2: τ3: 1:1:1 (in Nm); time: 0.5sec)

Kinematic modelling of three link robot manipulator and joint torque optimization … (Neeta Sahay)
164  ISSN: 2722-2586

Angular Rotation Vs. Applied Torque

Angular Rotation (rad)

3
2
theta1
1
theta2
0
theta3
-1 0 5 10 15
Applied Torque (Nm)

Figure 3. Angular position versus applied torque (τ1: τ2: τ3: 2.25:1.5:1 (in Nm); time: 0.5sec)

Angular Rotation Vs. Applied Torque

Angular Rotation (rad)

2
theta1
1
theta2
0 theta3
0 5 10 15 20 25
-1
Applied Torque (Nm)

Figure 4. Angular position versus applied torque (τ1: τ2: τ3: 4:2:1 (in Nm); time: 0.5sec)

Angular Rotation Vs. Applied Torque

Angular Rotation (rad)

4
2
theta1
0
theta2
-2 0 10 20 30 40
theta3
-4
Applied Torque (Nm)

Figure 5. Angular position versus applied torque (τ1: τ2: τ3: 6.25:2.5:1 (in Nm); time: 0.5sec)

Angular Rotation Vs. Applied Torque

3
Angular Rotation (rad)

2
1
0 theta1
-1 0 10 20 30 40 50 theta2
-2
theta3
-3
-4
Applied Torque (Nm)

Figure 6. Angular position versus applied torque (τ1: τ2: τ3: 9:3:1 (in Nm); time: 0.5sec)

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 160-167
IAES Int J Rob & Autom ISSN: 2722-2586  165

Angular Rotation Vs. Applied Torque

6

Angular Rotation (rad)

4
2 Theta1
0 Theta2
-2 0 20 40 60 80 100
Theta3
-4
Applied Torque (Nm)

Figure 7. Angular position versus applied torque (τ1: τ2: τ3: 16:4:1 (in Nm); time: 0.5sec)

Angular Rotation Vs. Applied Torque

15
Angular Rotation (rad)

10
5 Theta1
0 Theta2
0 50 100 150
-5 Theta3
-10
Applied Torque (Nm)

Figure 8. Angular position versus applied torque (τ1: τ2: τ3: 25:5:1 (in Nm); time: 0.5sec)

5. TORQUE OPTIMIZATION USING GENETIC ALGORITHM IN MATLAB

To optimize τ3, joint torque in the third joint actuator has been plotted against τ 2 and 𝜃2 as shown in
Figure 9. Genetic algorithm has been applied to minimize the objective function given in (19). Optimized
value of 𝜽𝟑 has been evaluated by plotting 𝜽𝟑 with respect to evaluated 𝜽𝟏 and 𝜽𝟐 as shown in Figure 10. The
optimized results of all the required parameters are given in Table 1.

Figure 9. Surface generation of T3 for torque Figure 10. Surface generation of θ3 with respect to θ1
optimization and θ2

Table 1. Optimized torque and corresponding angular displacement

τ1(in Nm) τ2(in Nm) τ3(in Nm) θ1 (rad) θ2(rad) θ3(rad)
31.25 12.5 1.3658 1.1376 -0.1861 1.3202
Kinematic modelling of three link robot manipulator and joint torque optimization … (Neeta Sahay)
166  ISSN: 2722-2586
2
𝜏3 = 𝑎 + 𝑏𝑠𝑖𝑛(𝑚𝜋𝑥1 𝑥2 ) + 𝑐𝑒 −(𝑤𝑥2 ) (19)

where a, b, c, m, and w are constant coefficients of the objective function (19) and evaluated in MATLAB as

𝑎 = 2.683
𝑏 = 0.9224
𝑐 = −0.4006
𝑚 = 0.2138
𝑤 = −0.6463

𝑥1 𝑎𝑛𝑑 𝑥2 are the torque required and resulting angular displacement of the link2, respectively. Boundary
condition has been placed as 1 ≤ 𝑥1 ≤ 12.5 and −3 ≤ 𝑥2 ≤ 0.
From Table 1, it is evident that the results show a successful optimization using genetic algorithm on
𝝅
the problem of torque optimization. For practical implementation, assumptions can be made that 𝟎 ≤ 𝜽𝟏 ≤ ,
𝟐
𝝅 𝝅
𝟎 ≤ 𝜽𝟐 ≤ and 𝟎 ≤ 𝜽𝟑 ≤ and the condition|𝜃2 | ≤ 2𝜃1 must be followed if θ1 is positive and θ2 found to be
𝟐 𝟐
negative. Therefore, the result shown is to satisfy the above-mentioned condition and the joint torques as well
as corresponding angular displacement may be evaluated as given in Table 1.

6. CONCLUSION
The manipulator dynamics, as stated in (2) to (4), represent the non-linearity of the robot arm
system. Here the open chain of links and its corresponding behavior of the manipulator has been studied
which shows the angular position vs. applied torque characteristics. The optimum value of torque with
respect to the desired movement of the manipulator arm has been evaluated using a genetic algorithm with
boundary conditions in MATLAB which also satisfies the relative angular position constraints. Also, the
system may be assumed to be a series of three inverted pendulums exhibiting the nature of a chaotic system
which is certainly nonlinear in nature. Therefore, stability of the system may be achieved with the help of a
sliding mode controller as well as by behavior-based control where it is to split a complex dynamic into
several simple equations which are quietly related to the problem stated in this research paper.

ACKNOWLEDGEMENTS
This research received no specific grant from any funding agency in the public, commercial, or
not-for-profit sectors.

REFERENCES
[1] D. P. Garg and M. Kumar, “Optimization techniques applied to multiple manipulators for path planning and torque minimization,”
Engineering Applications of Artificial Intelligence, vol. 15, no. 3–4, pp. 241–252, Jun. 2002, doi: 10.1016/S0952-1976(02)00067-2.
[2] S. Parsa, C. M. Saaj, H. R. Daniali, and R. Ghaderi, On-orbit servicing: Novel algorithms for motion control o robot
manipulators. 2008.
[3] H. Korayem and M. Irani, “Maximum dynamic load determination of mobile manipulators via nonlinear optimal feedback,”
Scientia Iranica, vol. 17, no. 2 B. pp. 121–135, Mar. 2010.
[4] M. Aghanouri, A. Habibollahi, A. Esmaeili, H. Faghihian, and M. Koloushani, “Optimization of robotic manipulators parameters
modeled with integrated equations of actuators and links,” in Electrodynamic and Mechatronic Systems - Proceedings of 2011,
3rd International Students Conference on Electrodynamics and Mechatronics, SCE III, Oct. 2011, pp. 31–36. doi:
10.1109/SCE.2011.6092120.
[5] V. Gupta, H. Chaudhary, and S. K. Saha, “Dynamics and actuating torque optimization of planar robots,” Journal of Mechanical
Science and Technology, vol. 29, no. 7, pp. 2699–2704, Jul. 2015, doi: 10.1007/s12206-015-0517-z.
[6] M. Taheri, “Nonlinear dynamic modeling and optimal motion analysis of two-link manipulators,” IAES International Journal of
Robotics and Automation (IJRA), vol. 5, no. 1, p. 61, Mar. 2016, doi: 10.11591/ijra.v5i1.pp61-66.
[7] R. Qi, A. Khajepour, and W. W. Melek, “Redundancy resolution and disturbance rejection via torque optimization in hybrid
cable-driven robots,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 52, no. 7, pp. 4069–4079, Jul. 2022,
doi: 10.1109/TSMC.2021.3091653.
[8] S. Baressi Šegota, N. Anđelić, I. Lorencin, M. Saga, and Z. Car, “Path planning optimization of six-degree-of-freedom robotic
manipulators using evolutionary algorithms,” International Journal of Advanced Robotic Systems, vol. 17, no. 2, p.
172988142090807, Mar. 2020, doi: 10.1177/1729881420908076.
[9] A. Televnoy, S. E. Ivanov, T. Zudilova, and L. N. Ivanova, “Transformation method for a nonlinear manipulator model,”
Procedia Computer Science, vol. 193, pp. 295–305, 2021, doi: 10.1016/j.procs.2021.10.030.
[10] I. Agustian, N. Daratha, R. Faurina, A. Suandi, and S. Sulistyaningsih, “Robot manipulator control with inverse kinematics PD-
pseudoinverse jacobian and forward kinematics denavit hartenberg,” Jurnal Elektronika dan Telekomunikasi, vol. 21, no. 1, p. 8,
Aug. 2021, doi: 10.14203/jet.v21.8-18.
[11] H. N. Rahimi and M. Nazemizadeh, “Dynamic analysis and intelligent control techniques for flexible manipulators: a review,”
Advanced Robotics, vol. 28, no. 2, pp. 63–76, Jan. 2014, doi: 10.1080/01691864.2013.839079.

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 160-167
IAES Int J Rob & Autom ISSN: 2722-2586  167

[12] M. H. Korayem, H. N. Rahimi, A. Nikoobin, and M. Nazemizadeh, “Maximum allowable dynamic payload for flexible mobile
robotic manipulators,” Latin American Applied Research, vol. 43, no. 1, pp. 29–35, Jan. 2013.
[13] J. Wu, J. Qiu, and H. Ye, “Torque optimization method of a 3-DOF redundant parallel manipulator based on actuator torque
range,” Journal of Mechanisms and Robotics, vol. 15, no. 2, Jun. 2023, doi: 10.1115/1.4054618.
[14] W. S. Tang and J. Wang, “Two recurrent neural networks for local joint torque optimization of kinematically redundant
manipulators,” IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, vol. 30, no. 1, pp. 120–128, 2000, doi:
10.1109/3477.826952.
[15] Y. Zhang and J. Wang, “A dual neural network for constrained joint torque optimization of kinematically redundant
manipulators,” IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, vol. 32, no. 5, pp. 654–662, Oct.
2002, doi: 10.1109/TSMCB.2002.1033184.
[16] X. Gao, M.-T. Du, L. Zhai, Y. Zhang, H.-X. Sun, and X.-P. Diao, “Research on joint torque optimization method of redundant
space manipulators with vibration suppression,” ITM Web of Conferences, vol. 12, p. 1004, 2017, doi:
10.1051/itmconf/20171201004.
[17] M. Naghshineh and M. Keshmiri, “Control and real time optimization of overactuated cooperative manipulators,” IEEE
Conference Number, no. August, pp. 953–960, 2003.
[18] A. Wolniakowski, C. Valsamos, K. Miatliuk, V. Moulianitis, and N. Aspragathos, “Optimization of dynamic task location within
a manipulator’s workspace for the utilization of the minimum required joint torques,” Electronics (Switzerland), vol. 10, no. 3, pp.
1–18, Jan. 2021, doi: 10.3390/electronics10030288.
[19] J. Woolfrey, W. Lu, and D. Liu, “A control method for joint torque minimization of redundant manipulators handling large
external forces,” Journal of Intelligent and Robotic Systems: Theory and Applications, vol. 96, no. 1, pp. 3–16, Jan. 2019, doi:
10.1007/s10846-018-0964-8.
[20] E. C. Agbaraji, I. Obiora-Dimson, and H. C. Inyiama, “Joint torque and motion computational analysis for robotic manipulator
arm design,” Journal of Engineering and Applied Sciences, vol. 12, pp. 1–9, Jan. 2018.
[21] H. Singh, N. Dhillon, and I. Ansari, “Forward and inverse kinematics solution for six DOF with the help of robotics tool box in
matlab,” International Journal of Application or Innovation in Engineering & Management, vol. 4, no. 3, pp. 17–22, 2015.
[22] M. Brandstotter, A. Angerer, and M. Hofbaur, “An analytical solution of the inverse kinematics problem of industrial serial
manipulators with an ortho-parallel basis and spherical wrist,” in Proceedings of the Austrian Robotics Workshop 2014, 2014, pp.
1–11.
[23] V. I. Petrenko, F. B. Tebueva, M. M. Gyrchinsky, V. O. Antonov, and J. A. Shutova, “The method of forming a geometric
solution of the inverse kinematics problem for chains with kinematic pairs of rotational type only,” IOP Conference Series:
Materials Science and Engineering, vol. 450, pp. 1–6, Nov. 2018, doi: 10.1088/1757-899X/450/4/042016.
[24] J.-D. Sun, G.-Z. Cao, W.-B. Li, Y.-X. Liang, and S.-D. Huang, “Analytical inverse kinematic solution using the D-H method for a
6-DOF robot,” in 2017 14th International Conference on Ubiquitous Robots and Ambient Intelligence (URAI), Jun. 2017, pp.
714–716. doi: 10.1109/URAI.2017.7992807.
[25] P. Badoniya and J. George, “Two link planar robot manipulator mechanism analysis with MATLAB,” International Journal for
Research in Applied Science and Engineering Technology (IJRASET), vol. 6, no. 7, pp. 778–788, 2018.

BIOGRAPHIES OF AUTHORS

Neeta Sahay received her B.Tech. degree in Electronics and Instrumentation

Engineering from the University of Kalyani also received M.Tech. degree in Mechatronics
Engineering from NITTTR, Kolkata [MHRD]. She has more than 15 years of teaching
experience. She started her professional career as an asst. professor in HIT, Haldia (Under
MAKAUT, WB) and then joined AOT, Adisaptagram as an asst. professor. Presently she is an
assistant professor at the Institute of Engineering & Management, Kolkata, West Bengal,
India, and pursuing a Ph.D. from NITTTR, Kolkata in the research area of Mechatronics and
Robotics. She can be contacted at [email protected].

Subrata Chattopadhyay was born in India in 1965. He received his Ph.D.

(Tech) in Instrumentation Engineering from the University of Calcutta, India in 2006, preceded
by M.Tech. [Instrumentation], B. Tech. [Electrical] and B. Sc (Hons) in Physics, in 1993,
1991, and 1987, respectively. He served as a deputy manager [Projects and Maintenance] in
Electrical and Instrumentation Engineering of Chemical and Manufacturing Industries in India
and then joined as an assistant professor in the Electrical Engineering Department of the
National Institute of Technical Teachers’ Training and Research, Kolkata, under the Ministry
of Education, Government of India in 2003. At present he is working as a professor in
Electrical Engineering and is in charge of NITTTR Kolkata Extension Centre, Bhubaneswar,
India. He introduced, as head of the Electrical Engineering Department, a new Post Graduate
Programme [M. Tech. in Mechatronics Engineering], the first of its kind in Eastern India at
NITTTR Kolkata, with the required development of the Department to accommodate the
same. He is highly involved in teaching and research and his present investigation is on the
innovation of noble techniques of measurement and control, based on sensor and transducer
development, process automation, PLC and distributed control systems, mechatronics, robotics,
bio-medical instrumentation, etc. He has guided two research scholars and at present, three
scholars are working under him for the Ph. D. degree. He has around 100 papers in international
journals and conference proceedings. He can be contacted at [email protected].

Kinematic modelling of three link robot manipulator and joint torque optimization … (Neeta Sahay)