See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/282382464

Design of PID controllers for 4-DOF planar and spatial manipulators

Article · April 2015

DOI: 10.1109/RACE.2015.7097269

CITATIONS READS

19 2,285

2 authors:

Ravi kumar Mandava Pandu Ranga Vundavilli

Indian Institute of Information Technology, Design & Manufacturing,Kurnool Indian Institute of Technology Bhubaneswar
63 PUBLICATIONS 537 CITATIONS 179 PUBLICATIONS 2,365 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Ravi kumar Mandava on 18 October 2017.

The user has requested enhancement of the downloaded file.

 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

Design of PID Controllers for 4-DOF Planar and

Spatial Manipulators
Ravi Kumar Mandava Pandu Ranga Vundavalli
Research scholar Assistant Professor
School of Mechanical sciences , IIT Bhubaneswar School of Mechanical sciences, IIT Bhubaneswar
Bhubaneswar, Odisha- 751013, India. Bhubaneswar, Odisha- 751013, India.
E-mail: [email protected] E-mail: [email protected]

Abstract- Robotic manipulators are the most important modelling of a spatial RRR flexible manipulator that
components in the present manufacturing industry for includes both link and joint flexibility. This algorithm was
handling materials and manipulation. It is important to note implemented by using the principle of virtual work, finite
that these manipulators are used in the complex planar and element method and recursive kinematic formulation. The
spatial work spaces to perform the task. Further, it is also very benefit of the algorithm was that it uses less set of equations
difficult to manipulate and dynamically control the that define the dynamics of flexible manipulator, which was
manipulators when the degrees of freedom (DOF) of the essential for the controller to reduce the computational cost.
manipulator are increased. In the present manuscript, the Moreover, the workspace of a robotic manipulator was
dynamic analysis and design of PID controller for two
determined after using an analytical method [6]. This
different manipulator configurations, namely planar and
spatial manipulators have been attempted. Moreover, both the
method was based on analytical criteria for determining the
manipulators considered in this study are provided with 4- singular behaviour of the mechanism. The authors
DOF and dynamic analysis has been carried out using attempted to compute the singularities after manipulating
Lagrange-Euler (L-E) formulation. The main difference the Jacobian of the robot condition. These singularities
between the planar and spatial manipulator is that the former were substituted in the wrist vector to acquire the range of
works in the two dimensional space and later works in a three motion of the robot wrist in 3D space.
dimensional space. In addition to the dynamic analysis, PID During manipulation of the robot, many problems need
controllers have also been designed for the said configurations
to be tracked simultaneously. Among those problems,
to follow a path between the given boundary conditions.
Further, the developed controllers are tested in simulations,
controlling of the robotic manipulator is a challenging task.
and found working satisfactorily. The basic idea of a robot controller is to make the joints of
a manipulator to track a desired trajectory to perform a
Keywords -Planar manipulator, spatial manipulator, dynamic specified task. The dynamic modelling of a simple 2-DOF
analysis, PID controller. robotic manipulator, using PD controller with online
gravity compensation was developed by Dung et al. [7].
Further, Ohri et al. [8] designed PID controller and sliding
I. INTRODUCTION mode controller for desired trajectory, in the presence of
Robotic manipulators are used in aerospace, highly nonlinear coupled dynamics with high accuracy. It
manufacturing, automotive industry and medical was observed that, when the payload was changed, the
applications. These manipulators are working in hazardous, SMC produced better response compared to PID control
unpredictable and inhospitable circumstances, which are strategy. In [9], the authors developed force-torque control
difficult for the human operations to perform the task. It strategy for two degrees of freedom robotic manipulator.
consists of three models, namely mechanical, electrical and Two types of control strategies were developed to control
control modules. In the mechanical point of view, robot the slow part of system dynamics, i.e. impedance control
manipulators are collection of serial or parallel links which and force/torque control with gravity compensation.
are connected with the help of revolute and/or prismatic Moreover, a classical PID controller [10] was implemented
joints between the base and the end effector frame. Number for a six-link, six-joint flexible robotic manipulator model.
of researchers had developed several methods to identify The flexibility was given to the joints of the system model.
the factors influencing the performance and accuracy of the Although the researchers had developed various
robotic manipulators [1]. approaches for serial and parallel manipulators, those
The serial manipulator with 3-revolute joints that can strategies may be divided into two categories, i.e. joint-
operate in 2-D space was developed by soomro [2] and Fu space control and workspace control. The joint space
et al. [3]. In both the cases, the authors had used forward controller considers the system as an independent SISO
kinematics approach to establish the mathematical model of system to implement the developed joint space control
the manipulator. Further, these mathematical models were strategy. Where, the workspace control is more attractive
used to determine mass moment of inertia, velocity and than joint space control because the tasks, such as painting,
acceleration of various joints of the manipulator. Moreover, assembly etc., are performed by the end effectors in the
Hatem et al. [4] developed a kinematic and dynamic model robots workspace [11]. In addition to the above works, Su
of a 3-DOF planar and spatial manipulators to overcome the et al. [12] designed and explained a robust auto-disturbance
control and noise factors. In [5], a general purpose rejection controller for 6-DOF parallel manipulator in joint
algorithm had been developed for accurate dynamic space. Moreover, in [13], the authors discussed the

ISBN: 978−81−925974−3−0
 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

development of controller for robotic manipulators with length of the links and included angle between them. The
complex nonlinear dynamics. Once the controllers are homogeneous transformation matrices for 4-DOF planar
developed, the tuning of the gains of the controller plays an and spatial manipulators are given by equations (1) and (2),
important role in determining the performance of the respectively.
controller. There are different types of tuning methods Table 2: The D-H parameters of the 4-DOF spatial manipulator are as
developed by the researchers such as manual tuning, follows:
Ziegler and Nichols tuning and online tuning of the
Link (i) αi di θi
controllers. Ziegler and Nichols [14] proposed the first i

tuning method for proportional-integral-derivative (PID) 1 0 900 0 θ1

controller. In [15], the authors designed a dynamic model 2 L2 0 0 θ2
and PID controller for a 2-DOF planar manipulator by
using the manual tuning procedure. The present manuscript 3 L3 0 0 θ3
deals with the development of a PID controller for two 4 L4 0 0 θ4
robotic manipulators, namely 4-DOF planar and spatial
robots, respectively.
The paper is organized as follows: section 2 explains the
mathematical formulation of the problem, i.e. kinematics
and dynamics of the robotic manipulator. Section 3
describes the design of a PID controller for both the 4-DOF
planar and spatial manipulators. Section 4 shows the results
of the simulation. Finally, section 5 concludes the paper.

II. MATHEMATICAL FORMULATION OF THE

PROBLEM
The present section focused on the development of
forward kinematics and dynamic modelling of two four-link
robotic manipulators, such as planar (refer to Fig. 1) and Fig 1: 4 DOF planar manipulator.
spatial (refer to Fig. 2) manipulators. The links of the
manipulators are considered to have masses and lengths
equal to {m1, m2, m3, m4} and {L1, L2, L3, L4}, respectively.
Let the included angles made by the links (L1 though L4) of
the manipulators are given as θ1, θ2, θ3 and θ4 in order to
define a position for the end effector. A systematic method
known as D-H parameter setting has been used to assign the
co-ordinate frames for the joints of the manipulator to
derive the transformation matrix, that represent the forward
kinematic model for both the manipulators. It is important
to note that forward kinematic model helps in determining
the position and orientation of end effector with respect to
base. The D-H parameters obtained for 4-DOF planar and
spatial manipulators are given in tables 1 and 2,
respectively.
Table 1: The D-H parameters of the 4-DOF planar manipulator are as Fig 2: 4 DOF spatial manipulator.
follows:
Link(i) i αi di θi
0
1 L1 0 0 θ1 0
0 0 0 1
2 L2 0 0 θ2
------------------ (2)
3 L3 0 0 θ3
Where the notations are referring to C1=cos(θ1), S1=sin(θ1),
4 L4 0 0 θ4 C12=cos(θ1+θ2), S12=sin(θ1+θ2), C123=cos(θ1+θ2+θ3),
S123=sin(θ1+θ2+θ3), C1234=cos(θ1+θ2+θ3+θ4) and
0 S1234=sin(θ1+θ2+θ3+θ4).
0
0 Further, the dynamics of the robotic manipulators have
0 0 1 0
been derived after considering Lagrange-Euler formulation,
0 0 0 1
as the dynamics are going to play an important role in
------------------ (1) restricting the real time performance of the manipulator
while executing a task. It is important to note that these
This information is further used in determining the
restrictions are due to the velocity and acceleration of the
homogeneous transformation matrix (0T4) that tells the
masses that introduce the inertia of the moving masses into
relationship between end-effector and base, in terms of

ISBN: 978−81−925974−3−0
 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

the system. Therefore, it needs to be carefully modelled Kd=Derivative gain

during the dynamic modelling of the manipulator. The
dynamic equation of motion is considered as a combination Ki= Integral gain
of accelerations produced by inertia forces, velocity The expanded form of the equation (4) after including the
produced by centrifugal or coriolis components and gravity meaning of and are given in equation (5).
terms of the manipulator. The dynamic equation of motion,
that is obtained from Lagrange-Euler formulation to -------- (5)
calculate the torques required at different joints are given in
equation (3). =
Where the values of i is varying from 1 through n, which
represent the number of links of the manipulator.

, 1,2, … … … ----------------- (3)

Where = force applied at joint due to motion of links IV. RESULTS & DISCUSSIONS

= Displacement of the joint. The main aim of the present research is to design and
develop a PID control strategy for the control of 4-DOF
= Velocity of the joint j. planar and spatial manipulators. Once the dynamic models
and control strategies are developed, both the manipulator
= Acceleration of the joint j. models are tested in simulations to check the controlling
Where the inertia coefficient is is accuracy of the strategies. It is important to note that the
controller gain values have been tuned using manual tuning
∑ , , 1,2, … … … method for a set of boundary conditions. The three gain
values (Kp, Kd and Ki) that resulted in best control action
The Coriolis force coefficient is for 4-DOF planar and spatial manipulators are given in
tables 3 and 4, respectively. The boundary conditions
∑ , , , 1,2, … … … utilized for the four joints of the planar and spatial
manipulator are given in equation (6) and (7) respectively.
The gravity coefficient is
Table 3: The tuning parameters of the planar manipulator are as follows:
∑ , 1,2, … … … JOINTS Kp Ki Kd
1 110 40 90
III. DESIGN OF PID CONTROLLER 2 90 35 80

The purpose of using PID controller in controlling the 3 80 30 70

robotic manipulator is to position the end-effector at the 4 70 25 60
desired location after following the boundary conditions. It
is important to note that the design of PID controller is
relatively simple, when the transmission mechanism Table 4: The tuning parameters of the spatial manipulator are as follows:
includes gears. The PID control algorithm works on
manipulation of the controller gains (Kp-gain of JOINTS Kp Ki Kd
proportional controller, Ki -gain of integral controller, Kd - 1 100 35 90
gain of derivative controller) obtained based on the
2 90 35 75
magnitude of error signal. The proportional part of the PID
controller is obtained by multiplying Kp with error signal 3 80 30 65
and it will reduce large part of the overall error. Generally, 4 70 30 55
PD or PID controllers are used in the closed loop system to
locate the end effector at the exact location. The integral The boundary conditions for planar manipulator:
part of the controller gain is obtained by multiplying the
integral value of the error with KI time. Finally derivative 450 < < 900
part of the controller gain is obtained by multiplying the 300 < < 600
first derivative of the error with derivative constant KD. It is
to be noted that the importance of KD is to counteract the -450 < < 900
influence of KP and KI terms when the output changes -300 < < 900 --------------- (6)
quickly. Further, the derivative term also helps in reducing
the overshoot and ringing and it is not going to exhibit any The boundary conditions for spatial manipulator:
effect on final error in position. The expression for joint- 00 < < 1800
based PID controller implemented in the study is given in
equation (4). 300 < < -450
--------------- (4) -450 < < 900
Where Kp= Proportional gain -300 < < 450 --------------- (7)

ISBN: 978−81−925974−3−0
 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

(a) Error on planar manipulator (a)

(b)
(b) Error on spatial manipulator
Fig 3:Schematic diagram showing the variation of error at joint 1, joint 2,
joint 3, joint 4.

The variations in the value of error in angular position

of all the four joints for both the planar and spatial
manipulators are shown in Fig. 3(a) and (b), respectively. It
has been observed that the value of error for joints 1 and 2
are seen to be low when compared with joints 3 and 4 for
both the manipulators. It may be due to the fact that the
angular distance covered by joints 1 and 2 are less than that
of the joints 3 and 4 for the same interval of time in order to
cover the segment of travel associated with that joint. It is
interesting to note that all the joints are settled by 4 seconds
and this result can be comparable with the existing
(c)
literature [8, 15]. Further the variation of torques required at
various joints, of the 4-DOF planar manipulator is shown in
Fig. 4(a), (b), (c) and (d) respectively. From the above
figures, it can be observed that the torque required at joint 1
is more than that of the joints 2, 3 and 4. This may be due to
the reason, that the link 1 is driving the links 2, 3 and 4.
Therefore, joint 1 requires more torque when compared to
other joints. Similarlly, the variation of torques for various
joints of 4-DOF spatial manipulator is shown in Fig. (5). It
is interesting to note that similar findings as discussed for
planar manipulator are also true for the spatial manipulator.
Moreover, the coverage of work volume for 4-DOF
planar and spatial manipulators for the given boundary
conditions are shown in Fig. 6(a) and (b), respectively. It
has been observed that both the manipulators have covered (d)
the work space smoothly after utilizing the developed PID
Fig 4:Schematic diagram showing the variation of torque in planar
controller. manipulator at (a) joint 1, (b) joint 2, (c) joint 3, (d) joint 4.

ISBN: 978−81−925974−3−0
 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

(a)

(a)

(b)

(b)
Fig 6:Schematic diagram showing the variation of work volume (a) Planar
manipulator, (b) Spatial manipulator.

V. CONCLUSIONS
In the present paper, the forward kinematic modelling,
dynamic analysis and design of PID controllers for two 4-
DOF planar and spatial robots are developed successfully.
It is important to note that the manual method of tuning
(c) have the PID controllers been adopted to tune the
proportional, integral and derivative gains of the controllers
for both the robots. Once the controllers are developed, they
are tested in simulations to verify the performance of the
developed controllers. Both the PID controllers developed
to control the planar and spatial manipulators are found to
give satisfactory performance measures while reaching the
specified boundary conditions with reasonable torque
(d) requirements at various joints.

VI. REFERENCES
[1] Swevers, J., Verdonck, W., and De Schutter, J., “Dynamic Model
Identification for Industrial Robots,” IEEE Control Systems
Magazine,Vol. 27, pp. 58-71 (2007).
(d) [2] Zulfiqar Ali soomro., “Kinematic Modeling and Simulation of 2-D
Link and 3-R Pendulum Serial Manipulator Robotic Arm,”
Fig 5:Schematic diagram showing the variation of torque in spatial International Researcher, Vol.2,Issue No.4.December (2013).
manipulator at (a) joint 1, (b) joint 2, (c) joint 3, (d) joint 4. [3] Robotics: "Control, Sensing, Vision, and Intelligence," King Sun
Fu, Rafael C. González, C. S. George Lee,McGraw-Hill, (1987 ).

ISBN: 978−81−925974−3−0
 International Conference on Robotics, Automation, Control and Embedded Systems – RACE 2015
18-20 February 2015, Hindustan University, Chennai, India.

[4] Hatem A. Al-Dois., A.K Jha., and R.B Mishra., “Investigations into
the Parameters Influencing the Dynamic Performance of 3-RRR
Planar & Articulated Robot Manipulators,” Tamkang Journal of
Science and Engineering, Vol.14, No. 4, pp.313-322 (2011).
[5] Rajesh K Moolam., Francesco Braghin., Federico Vicentini.,
“Dynamic Modeling and Simulation of Spatial Manipulator with
Flexible Links and Joints,” 11th International Conference on
Vibration Problems Z. Dimitrovov´a et.al. (eds.) Lisbon, Portugal, 9–
12 September(2013).
[6] Khusheep Goyal, Davinder Sethi, "An Analytical Method to Find
Workspace of a Robotic Manipulator,"Journal of mechanical
engineering.Vol. ME 41, No.1, June(2010).
[7] Le Tien Dung., Hee-Jun Kang., and Young-shick Ro., “Robot
Manipulator Modeling in Matlab Simmechanics with PD Control
and Online Gravity Compensation,” IFOST 2010 Proceedings, IEEE
(2010).
[8] Jyothi Ohri, Dhaval R. Vyas, Pretty Neelam Topno, "Comparison of
Robustness of PID Control and Sliding Mode Control of Robotic
Manipulator," International symposium on devices MEMS,
Intelligent systems & Communication(ISDMISC), (2011).
[9] Victor Titov., I.Shardyko., Sergey Isaenko., “Force-Torque Control
Implementation for 2 DoF Manipulator”, 24th DAAAM International
Symposium on Intelligent Manufacturing and Automation (2013).
[10] Prasenjit Sarkhel., Nilotpal Banerjee., Nirmal Baran Hui., “Analysis
and Control of a Six Link Serial Manipulator with Flexible Joints”
142-inacomm(2013).
[11] Raouf Fareh, Mohamad Saad, and Maarouf Saad, "Workspace
Tracking Trajectory for 7-DOF ANAT Robot Using a Hierarchical
Control Strategy" 20th Mediterranean Conference on Control &
Automation (MED) Barcelona, Spain (2012).
[12] Y. X. Su, B. Y. Duan, C. H. Zheng, Y. F. Zhang, G. D. Chen, and J.
W. Mi, “Disturbance-rejection High-precision Motion Control of a
Stewart Platform,” IEEE Trans. Contr. Syst. Technol., vol. 12, no. 3,
pp. 364-374, (2004).
[13] Lewis, F.L. & et al. "Robotics Mechanical Engineer Handbook". Ed.
Frank Kreith. Boca Raton: CRC Press LLC (1999).
[14] J. G. Ziegler, and N.B. Nichols, "Optimum Setting for Automatic
Controllers," Trans. ASME, vol. 64, pp. 759-768, (1942).
[15] David I. Robles G., “PID Control Dynamics of a Robotic Arm
Manipulator with Two Degrees of Freedom,” August 17, (2012).

ISBN: 978−81−925974−3−0

View publication stats