Speed Control System in Mobile Robot Based On Bezi

IOP Conference Series: Materials Science and Engineering
PAPER • OPEN ACCESS
Speed control system in mobile robot based on Bezier curve trajectory

To cite this article: M F Alfatih et al 2021 IOP Conf. Ser.: Mater. Sci. Eng. 1108 012015
View the article online for updates and enhancements.
This content was downloaded from IP address 191.101.77.247 on 24/03/2021 at 02:04

International Conference on Innovation in Science and Technology (ICIST 2019) IOP Publishing
IOP Conf. Series: Materials Science and Engineering 1108 (2021) 012015 doi:10.1088/1757-899X/1108/1/012015
Speed control system in mobile robot based on Bezier curve

trajectory
M F Alfatih*, M A Riyadi and I Setiawan

Department of Electrical Engineering, Faculty of Engineering, Universitas
Diponegoro, Semarang, Indonesia
*[email protected]
Abstract. The development of robot technology is currently growing rapidly. Mobile robot is
one robot with many applications in everyday life. One of the important things in the field of
mobile robots is the problem of trajectories or trajectories, in this case how robots can track
approved paths. In this study, the mobile robot control system has been tested to track tracks.
The path used in this study is the third-order Bezier Order, with control points and end points of
the predetermined pathway. Based on the results of simulation tests on the development of
mobile robot trajectories, based on the results of simulation tests on mobile robots, it is found
that the performance of the track tracking control system is very dependent on the placement of
the Bezier curve control points or in other words the actual path formed by the robot is greatly
influenced by the shape of the Bezier trajectory that was designed.
1. Introduction
Robot is a mechanical device that can perform physical tasks [1], either using human supervision and
control, using pre-defined programs or utilizing the principles of artificial intelligence [2]. Research on
mobile robots has been intensively discussed [3–5]. Furthermore, many mobile robots have been used
for various purposes such as industry [6], defense, places that have not been touched by humans (e.g.
the Mars Rover Robot is used for planetary exploration, Mars) [7], or even for educational purposes,
such as robots educative and also likes things like robot ball competition, grass cutting robot
competition, fire source search robot competition [8].
One of the main tasks that must be completed by each mobile robot is to get to the expected locations
and find out their position and orientation at all times (navigation system) [9]. A simple technique that
can be used is by moving the robot to follow a straight path to the destination to rotate in place according
to the expected orientation [10]. Because of the simplicity of control, in general this method is widely
used in practical robots, such as those used in nursing robots (nursing robots) designed by Johann
Borentein [11]. One of the real disadvantages of using this technique is that robots cannot move
smoothly for points that must be passed. So as to avoid this, the robot trajectory designed must be
continuous and smooth.
Bezier curves are commonly used in the field of computer graphics and in some literature have also
been successfully implemented as robot trajectory [12,13]. Bezier curves can be used to optimize robot
paths that are calculated with respect to length and curvature [14]. For robots to follow lane, it is
important for the speed of the left and right wheel drive angle to adjust gently. The Bezier curve method
can be applied to create curved lines to implementation continuous angular velocity [15].
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The drive system basically functions to move to drive the robot from one location to another. In this
case there are two types of driving systems on popular mobile robots: Ackerman driving which is
commonly used in automobile vehicles and differential driving (often found in wheelchairs with
automatic drive).
Based on previous research on mobile robots, there are still shortcomings in the form of reference
tracks that are still inaccurate in providing reference links to mobile robots. This paper will present a
construction of trajectories for mobile robots based on the Bezier curve method, with differential robot
type mobile robots which generally have a degree of flexibility in maneuvering or rotating in their
movements.
2. Methods
2.1. Mobile robot kinematics model

One type of mobile robot that is commonly used, especially for indoor operation is a mobile robot with
a steering or differential drive system (differential drive) [16]. The main reason is because it is relatively
more flexible in maneuvering and ease of control.
This type of robot basically has two main wheels, each of which is driven by a separate drive
(generally in the form of a permanent magnet DC motor with gear-reducing which serves to strengthen
the motor torque), as a counterweight generally this robot is also equipped with one or two castor wheels
which is placed at the back of the robot. Figure 1 shows the robot architecture seen from the top: If the
two cogs are rotating at the same speed, the robot will move in a straight direction, whereas if the speed
of one of the wheels is slower than the robot will move to form a curve with the direction of the path to
one the slower-moving wheels.
{A}
yA
C
VL
R VY

y VX
VR
L
r
x xA
Figure 1. Mobile robot position and orientation in Cartesian coordinate systems [16].
By its nature, this type of mobile robot is included in the category of non-holonomic robot: The position
and direction (orientation) are in a coupled state, in this case the robot cannot practically translate to the
left (or right) without first changing the robot's orientation.
Based on Figure 1 for the length of the wheel radius r, as well as the rotation speed of each right
wheel ωR and left wheel ωL respectively, then the linear speed of the right wheel 𝑣𝑅 (𝑡)and the left
wheel 𝑣𝐿 (𝑡)can be found using equation ( 1) and (2) following [17]:
vR (t )  r R (t ) (1)
vL (t )  rL (t )
(2)
2
When the robot rotates for a moment with the length of the radius R measured from the center of
rotation and the center of the two wheels (Figure 3 can be used as a reference), the rotation speed at each
point of the robot will always be the same (the robot is a rigid mechanical system), so the following
equations (3) and/or (4) apply to calculate the rotational speed of the robot.
vR (3)
 (t ) 
RL
2
vL (4)
 (t ) 
RL
2
Based on equations (3) and (4) the rotation speed of the robot can be calculated only based on
information from the two linear wheels’ speeds of the robot:
vR (t )  vL (t ) (5)
 (t ) 
L
Whereas the radius of the path can be searched by substituting equation (5) into equation (6), and solving
it for R:
L (vR  vL )
R (6)
2 (vR  vL )
Seen from equation (6), the radius of the momentary circle trajectory is inversely proportional to the
difference in the speed of the two robot wheels, the smaller the difference between the two wheel speeds,
the radius of the momentary circle formed by the robot trajectory is getting longer, and vice versa,
whereas if vR = vL, then R =∞R, or practically the robot will move to form a straight path. For the robot
to rotate at the center of its axis (R = 0), based on equation (6), the speed of the two wheels must be the
same in the opposite direction or vR = -vL
Based on equations (5) and (6), the linear velocity of the robot can be calculated using the following
equation (7):
vR (t )  vL (t ) (7)
v(t ) 
2
For simplicity, equations (6) and (7) can be collected in the form of matrix-vector equations as shown
in equation (8) below:
 v(t )   1 2 1  v (t )
(8)
 (t )   1
2  R
 
   L  1 L  vL (t ) 
Equation (8) above shows the direct kinematics relation between the linear velocity of the robot wheels
and the linear and angular speed of the robot, while equation (9) below shows the opposite relation:
vR (t ) 1 L 2   v(t )  (9)

v (t )     
 L  1  L 2   (t )
By knowing the linear and angular velocity of this robot at any time, then the speed at each Cartesian
axis can be found by projecting the robot velocity vectors to these axes, this is shown in equation (10)
below:
 x (t )  cos( (t )) 0
 y (t )   sin( (t )) 0  v(t )  (10)
     (t )
(t )   0 1  
From equation (10) it can be seen that for each moment the position and orientation of the robot in the
Cartesian coordinate system can be calculated by integrating each equation term:
3
t
x(t )   v( ) cos( ( ))d (11)
0
t
y (t )   v( ) sin( ( ))d (12)
0
t
 (t )    ( )d (13)
0
The results of solving the equations above are given below:
(vR (t )  vL (t ))t (14)

 (t )   0
L
x(t )  x0  sin( (t ))   0 )  sin( 0 )

v(t )
(15)
 (t )
y (t )  y0 
v(t )
cos( (t ))   0 )  cos( 0 ) (16)
 (t )
2.2. Motion control

As mentioned earlier, the main purpose of a robot control system is to move the robot from its initial
state (P0) to its final location (P3) by following the expected trajectory. This is illustrated in Figure 2:
Figure 2. Illustration of robot trajectory.
To achieve these objectives a motion control system is needed. At the lowest level, this system basically
regulates the rotational speed of each robot's wheels in accordance with the Reference as shown in Figure
3:
Figure 3. Block diagram control of DC motor in robot mobile.
As discussed earlier, the feedback control system that will be used to control the DC motor is the PID
control, with the equation of the transfer function.
4
𝑡 𝑑𝑒(𝑡)
𝐶𝑂(𝑡) = 𝐾𝑃 𝑒(𝑡) + 𝐾𝐼 ∫0 𝑒(𝜏)𝑑𝜏 + 𝐾𝐷 (17)
𝑑𝑡
With 𝐶𝑂(𝑡)= Output controller

Kp = Proportional constant
Ti = time integral
𝑇𝐷 = time derivative
𝐾𝐼 = gain integral (𝐾𝑃 /𝑇𝐼 )
𝐾𝐷 = gain derivative (𝐾𝑃 𝑇𝐷 )
While the DC motor itself basically has the following order 1 transfer function:
b (18)
H (s) 
Ts  1
with
b = gain static DC motor
T = time constant DC motor
Figure 4. Flowchart simulation trajectory.
When the system starts or starts, the operator must input the values of X and Y starting position, end
position and control points and robot travel time. The system will calculate the robotic path, curvature
and rotational speed of the robot from the start point to the end point. After that the system will run the
Bezier curve trajectory generator algorithm to graph the robot's trajectory based on the input value from
the operator.
2.3. Simulation of the overall mobile robot system control block

In general, the control system model that was simulated is shown in Figure 5.
5
Figure 5. Robot control system block.
In this simulation the quantities of the PID control amplifier was searched empirically for the values of
the given DC motor parameters. The ultimate goal of this simulation is basically to look at the dynamics
of the robot's trajectory to reference the expected trajectory.
Bezier curves are parametric curves that are often used in graphical applications and related fields
[4,14,18–20]. This curve is determined by several control points and always the initial and final control
points are determined. The shape can be changed by moving the control point. The two-dimensional
Bezier curve of order n is represented in the following equation:
𝑛!
𝑃 (𝑡) = ∑𝑛𝑖=0 (𝑖!(𝑛−𝑖)!) 𝑡 𝑖 (1 − 𝑡)𝑛−𝑖 𝑃𝑖 , 𝑡𝜖[0,1] (19)
Where 𝑃(𝑡) = [𝑥 (𝑡), 𝑦(𝑡)]𝑇 and 𝑃𝑖 are two-dimensional Bézier curves and control points. Bezier curves
have geometrical properties that are useful for path planning:
 Each Bezier curve starts at 𝑃0 and ends at𝑃𝑛..
 The curve is tangent to the line connected by 𝑃0 and 𝑃1 and the line connected by 𝑃𝑛 and
𝑃𝑛−1 respectively at 𝑃0 and𝑃𝑛.:
𝑇(0) = 𝑛(𝑃1 − 𝑃0 ) (20)
𝑇(0) = 𝑛(𝑃𝑛 − 𝑃𝑛−1 ) (21)
 The second derivative in 𝑃0 must be determined by 𝑃0 , 𝑃1and𝑃2 ,and the second derivative in P_n
must be determined by 𝑃𝑛−2 and 𝑃𝑛−1 .
𝑆(0) = 𝑛(𝑛 − 1)(𝑃2 − 2𝑃1 + 𝑃0 ) (22)
𝑆(1) = 𝑛(𝑛 − 1)(𝑃5 − 2𝑃4 + 𝑃3 ) (23)

According to the above properties, the control points of the Bezier curve are calculated below. The
curvature at any point on the Bezier curve is determined by the first and second derivatives with respect
to the parameter 𝑢 as [21]:
𝑃(𝑞) = 𝑃0 (1 − 𝑞)3 + 𝑃1 𝑞(1 − 𝑞)2 + 3𝑃2 𝑞 2(1 − 𝑞) + 𝑃3 𝑞, 𝑞 ∈ [0.1] (24)
To achieve the path of continuous curvature, we mainly focus on ensuring continuity of curvature at the
waypoint connecting two curves.
3. Results and discussion

The application of the Bezier curve method for the trajectory of the mobile robot trajectory was tested
with different starting points (P0), control points 1 and 2 (P1, P2) and end points (P3). The observed
results are the accuracy of the track shape with the starting point (P0), control points 1 and 2 (P1, P2)
and end points (P3) entered, the robot's rotational speed profile, robot wheel speed and robot travel time
from starting point (P0) to end point (P3).
6
3.1. Simulation with starting point (𝑃1 = 0.0) and end point (𝑃3 = 400.400) with some varied control
points (𝑃1 , 𝑃2 ), and speed 10 cm/s
Figure 6. Bezier curve paths with P0 (0,0) and P3 (400,400) and velocity of 10 cm/s.
With reference to equation (24):
𝑃(𝑞) = 𝑃0 (1 − 𝑞)3 + 𝑃1 𝑞(1 − 𝑞)2 + 3𝑃2 𝑞 2(1 − 𝑞) + 𝑃3 𝑞, 𝑞 ∈ [0.1]
Figure 6 shows that the Bezier curve is formed with P0 (0,0), P3 (400,400) and P1 and P2 which are
varied and the robot speed is 10 cm / s. From this simulation the system that was built successfully
created a robot car reference path based on input parameters x and y. The results of tracing the actual
trace of the curve attempted experienced significant differences. As shown in the figure, the final
coordinates of each experiment differ. for the first experiment with P1 (150.0), p2 (250,400) detected
endpoints (350,410). for the second experiment with P1 (200.0), p2 (300,400) the endpoints detected
were (420,320), then for the third experiment with P1 (170.0), p2 (150,400) detected endpoints
(300,470).
Figure 7. Step response in the PID control scheme with the track reference in Figure 6.
In Figure 7, referring to the PID control equation in equation 17:
𝑡
𝑑𝑒(𝑡)
( ) ( )
𝐶𝑂 𝑡 = 𝐾𝑃 𝑒 𝑡 + 𝐾𝐼 ∫ 𝑒(𝜏)𝑑𝜏 + 𝐾𝐷
0 𝑑𝑡
And with equation 18 for a first order DC motor equation:
0.0274
𝐻 (𝑠) =
0.1 + 1
7
The response control system that is applied to the dynamics of the trajectory that is designed has a good
and fast level of lift.
3.2. Simulations on variations in parameters P3 with fixed speed and control points (P1, P2)
Figure 8. Bezier curve paths with variation parametersP3 with fixed speed and control points (P1, P2),
with Speeds of 10 cm/s.
Figure 8 shows the Bezier curve with the endpoint P3 that must be traveled by a robot that is located
differently. with the same control points, namely p1 (150.0), P2 (250,400) and P0 (0.0), and also the
robot's fixed speed is 10 cm / s. the graph shows that the reference path can be made accurately, but the
robot is not able to get to the final point correctly. for the first experiment with P3 (400,400) the robot
goes to P3 (348,448), for the second experiment with P3 (400,200) the robot goes to P3 (389,210) and
for the third experiment with P3 (300,200) the robot goes to P3 (210,151).
Figure 9 the response of the control system applied to the dynamics of the designed path has a stable
lifting rate, and the results are the same as from the first experiment.
8
3.3. Simulations with variations in speed, but with the same control points (P1 and P2) and end points
(P3)
Figure 10. Bezier curve paths with parameters P0 (0.0), P1 (150.0), P2 (250.400), P3 (400.400) and
with variation of speeds.
Figure 6 shows that the Bezier Bezier curves are formed with P0 (0,0), P1 (150,0), P2 (250,400), P3
(400,400) and different robot speeds. In this experiment the system that was built succeeded in making
the reference robot car line in accordance with the input parameters x and y. The results turned out to
have significant similarities, so the difference in speed of the robot did not affect the shape of the bezier
curve.
In Figure 11, the control system response that is applied to the dynamics of the designed path has a
smooth lift rate and different response times in the stability of the control system. For a speed of 10 cm
/ s with an amplitude of 0.08 at 0.4 seconds, then for a speed of 6 cm / s the output amplitude is 0.20 at
0.5 seconds, and for a speed of 3 cm / s the output amplitude is 0.50 at 0.6 seconds.
4. Conclusions
Based on testing and analysis conducted in this study, the performance of a track tracking mobile robot
control system is strongly influenced by the shape of the Bezier trajectory or trajectory given as a
reference. For the coordinates of the endpoint to which Xfinal, Yfinal, are, the relatively consistent control
points that give the final result with minimal errors are P1=Xfinal/2,0 and P2=Xfinal/2, Yfinal. For differential
9
type Mobile Robot, there is a restraint of the Y axis's final coordinate destination which is the maximum
Xfinal /2.
References
[1] Adriansyah A 2014 Perancangan Localization Menggunakan Metode Dead Reckoning Sinergi 18
25–30
[2] Bounini F, Gingras D, Pollart H and Gruyer D 2017 Modified artificial potential field method for
online path planning applications IEEE Intelligent Vehicles Symposium (IV) 180-185
[3] Yu W, Shuo W, Rui W and Min T 2017 Generation of temporal – spatial Bezier curve for
simultaneous arrival of multiple unmanned vehicles Inf. Sci. (Ny) 418–419 34–45
[4] Simba K R, Uchiyama N and Sano S 2013 Real-Time Trajectory Generation for Mobile Robots
in a Corridor-Like Space Using B ´ ezier Curves 37–41
[5] Dinçer Ü and Çevik M 2019 Improved trajectory planning of an industrial parallel mechanism
by a composite polynomial consisting of Bézier curves and cubic polynomials 132 248–263
[6] Li H and Savkin A V 2018 An algorithm for safe navigation of mobile robots by a sensor network
in dynamic cluttered industrial environments Robot. Comput. Integr. Manuf. 54 65–82
[7] Yang G J and Choi B W 2013 Implementation of joint space trajectory planning using RTO swith
considering velocity constraints for mobile robots IEEE ISR 1-3
[8] Xianglin Y, Zhiwen Z, Junhao X and Zhiqiang Z 2015 Trajectory Planning for RoboCup MSL
Mobile Robots Based on Bézier Curve and Voronoi Diagram 2552–2557
[9] Dixon J and Henlich O 1997 Mobile Robots Navigation
[10] Illah R 2004 Autonomous Mobile Robots (London, England: MIT Press)
[11] Borenstein J and Koren Y 1985 A Mobile Platform for Nursing Robots IEEE Trans. Ind. Electron.
IE-32 2 158–165
[12] Huang J, Farritor S M, Qadi A and Goddard S 2006 Localization and Follow-the-Leader Control
of a Heterogeneous Group of Mobile Robots 11 2 205–215
[13] Hwang J H, Arkin R C and Kwon D S 2003 Mobile robots at your fingertip: Bezier curve on-line
trajectory generation for supervisory control IEEE Int. Conf. Intell. Robot. Syst. 2 1444–1449
[14] Costanzi R, Fanelli F, Meli E, Ridolfi A and Allotta B 2016 Mobile Robots Based on Bezier
Curves IFAC-PapersOnLine 49 15 145–150
[15] Ishikawa H, Noguchi K, Maki R and Naitoh H 2009 Path generation with clothoid curve using
image processing for two-wheel-drive autonomous mobile robots 42 16
[16] Koren Y and Borenstein J 1991 Potential field methods and their inherent limitations for mobile
robot navigation Proc. - IEEE Int. Conf. Robot. Autom. 2 1398–1404
[17] Jusuf Dwi Kurnianto N H, Ali H A and Fernando A 2012 Navigasi Mobile Robot Berbasis
Trajektori dan Odometry dengan Pemulihan Jalur Secara Otomatis 66 37–39
[18] Xu L, Wang D, Song B and Cao M 2017 Global smooth path planning for mobile robots based
on continuous Bezier curve Proc. - 2017 Chinese Autom. Congr. CAC 2081–2085
[19] Zhou F, Song B and Tian G 2017 Bezier Curve Based Smooth Path Planning for Mobile Robot
[20] Hassani V and Lande S V 2018 Path Planning for Marine Vehicles using Bézier Curves IFAC-
PapersOnLine 51 29 305–310
[21] Zhang L, Sun L, Member I, Zhang S, Liu J and Member I 2015 Trajectory Planning for an Indoor
Mobile Robot Using Quintic Bezier Curves 757–762
10

Speed Control System in Mobile Robot Based On Bezi

Uploaded by

Copyright:

Available Formats

Speed Control System in Mobile Robot Based On Bezi

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Speed Control System in Mobile Robot Based On Bezi

Uploaded by

Copyright:

Available Formats

IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Speed control system in mobile robot based on Bezier curve trajectory

View the article online for updates and enhancements.

This content was downloaded from IP address 191.101.77.247 on 24/03/2021 at 02:04

Speed control system in mobile robot based on Bezier curve

M F Alfatih*, M A Riyadi and I Setiawan

2.1. Mobile robot kinematics model

vR (t ) 1 L 2   v(t )  (9)

The results of solving the equations above are given below:

(vR (t )  vL (t ))t (14)

x(t )  x0  sin( (t ))   0 )  sin( 0 )

2.2. Motion control

Figure 2. Illustration of robot trajectory.

Figure 3. Block diagram control of DC motor in robot mobile.

With 𝐶𝑂(𝑡)= Output controller

T = time constant DC motor

Figure 4. Flowchart simulation trajectory.

2.3. Simulation of the overall mobile robot system control block

Figure 5. Robot control system block.

𝑆(0) = 𝑛(𝑛 − 1)(𝑃2 − 2𝑃1 + 𝑃0 ) (22)

𝑆(1) = 𝑛(𝑛 − 1)(𝑃5 − 2𝑃4 + 𝑃3 ) (23)

𝑃(𝑞) = 𝑃0 (1 − 𝑞)3 + 𝑃1 𝑞(1 − 𝑞)2 + 3𝑃2 𝑞 2(1 − 𝑞) + 𝑃3 𝑞, 𝑞 ∈ [0.1] (24)

3. Results and discussion

You might also like