ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

A new Design and Control of a Two-Wheel Self-

Balancing Robot using the Arduino Microcontroller
[1]
S.Manikandan, [2] Reshma, [3] G Vani prasanna, [4] Ch Swetha, [5] V S Hariharan
[1]
Professor, [2][3][4] Student, [5] Principal
Dept of EEE, Balaji Institute of Technology and Science, Warangal, Telangana

Abstract: -- In the last decade, the open source community has expanded to make it possible for people to build complex products at
home. [1] In this thesis a two wheeled self-balancing robot has been designed. These types of robots can be based on the physical
problem of an inverted pendulum [2]. In this paper, we can see the design, construction and control of a two-wheel self-balancing
robot. This system consists of a pair of DC motor and an Arduino UNO R3 microcontroller board, make a robot which can balance
itself on two wheels the platform will not remain stable itself. Our job will be to balance the platform using distance sensors and to
maintain it horizontally. At first, we have decided to just balance the robot on its two wheels.

Keywords: — Robot, Arduino, AT mega 328, Control Systems, PID controller, Linear Quadratic Regulator.
This paper is arranged as follows: The Section 2 describes the
hardware and system architecture of the robot. The designs of
I. INTRODUCTION filters, inner control loop to equilibrate the two motors,and
To make a robot which can balance itself on two wheels. balancing control in Section 3. Finally Section 4 represents the
There will be only one axle connecting the two wheels and a experimental results, followed by some conclusion in Section
platform will be mounted on that .There will be a another 5.
platform above it. The platform will not remain stable itself.
Our job will be to balance the platform using distance sensors II. STRUCTURE OF THE TWO-WHEEL BALANCING
and to maintain it horizontal. At first we have decided to just ROBOT
balance the robot on its two wheels. Basically, a two-wheel
self-balancing robot is very similar to the inverted pendulum, In the structure of a self-balancing robot can be classified into
and which is an important test part in control system and three parts such as sensors, motor and motor control, and
research education purpose; let us see, for an example [5], [6]. develop board [4], [5]. In Section II-A introduces the
A Two-wheeled self-balancing vehicle commercially known application and advantage of the sensors on the proposed
as “Segway”.And also the Segway can never stay balancing robot, andalso how these sensors are employed to
upright.Besides the development of Segway, studies of two- obtain measurements of acceleration, and distance
wheel self-balancing robots have been widely reported. For traveled,androbot tilt angle. The Section II-B describes the
example, JOE [5] and nBot [6] are both early versions motor selection and control for the balancing robot. And
complete with inertia sensors and motor encoders and also Section II-C discusses the reason behind choosing the Arduino
along with on-vehicle microcontrollers. Arduino is an open develop board, and how it is deployed.
prototyping platform based up on Atmega processors .And It
will be a fast becoming popular platform for both education
[7] and product development, with applications ranging from
robotics [8], [9] to process control [10], [11] and networked
control [12].In this paper, we report a student project on the
basis of design, construction and control of a two-wheel self-
balancing robot with Ardunio software. The robot is driven by
two DC motors, and is equipped with an Arduino Uno board
which is based on the ATmega328 microprocessor, and also
including with a single-axis gyroscope and a 2-axis
accelerometer for attitude determination. In order to
compensate for gyro drifts common in COTS sensors, a
complementary filter is implemented [13]. For two wheel
control designs are based up on the linearized equations of
motion, such as a proportional-integral-differential (PID) Fig 1: Self Balancing Robot
control.

All Rights Reserved © 2018 IJEREEE 11

 ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

A. Selection and Application of Sensors UNO features 54 general-purpose digital I/O pins of which 15
In order to make balance of the robot, there are two things provide PWM output. And alsosome of the digital I/O pins
must needed one is robot’s tilt angle and angular velocity. It is also support serial communication such as I2C and SPI, as
possible by using MEMS sensors. And it may also including a well as interrupt handles. The board also features sixteen 10-
gyroscope, wheel-angle encoders, and alsoan accelerometer bit analog inputs for 0-5V input, giving a quantization limit of
and we can measure all the data needed for balancing control. 4.9mV.
1) Gyroscope: Gyroscope measures the angular rate around an
axes. Tilt angle can be obtained by integrating angular rate
over sampled time. An estimate of angular displacement is
obtained by integrating velocity signal over time.

2)Accelerometer: This meter is used to measures the total

external acceleration of the balancing robot, and also which
includes the gravitational and motion accelerations.
Accelerometer can measure the force of gravity and with that
information, the angle of robot can be obtained. Kalman filter Fig. 2. Architecture of the controlled system
is used for the fusion of outputs of two sensors. 3) Price and Expansions: As we all know Arduino boards are
3) Encoders: In generally the encoders return the rotation low-cost and expandable, where optional peripherals called
angles of individual motor shafts as digital signals, which are shields can be purchased as and when needed.
sent to the processor. Next, after conversion based on gear D. Power Supply : In generally power supply, the motors a
ratio and wheel radius, the distance traveled can be calculated. voltage between 12V to 16V, and for the development voltage
The encoder chosen for this project is the 64 cycle-per- between 5V to 15V.These analog voltages lie in therange from
revolution, which provides a resolution of 64 counts per 0V to 5V.
revolution of the motor shaft.
III. ANGLE ESTIMATION AND BALANCING
B. Motor Control Board CONTROL
One of the most important part in balancing robot is Dc
Motor. It emphasizes torque output instead of velocity, The fig 2 represents the system architecture of the self-
because it has to oppose the rotational moment that gravity balancing robot. A balancing controller which delivers a
applies on the robot.Hence, the motors need to provide enough motor-control signal, and a inner-loop controller for wheel
torque to correct the robot’s body pose back to a balanced synchronization. And also a Feedback is provided through a
position. complementary filter. The filter function is to provide an
Generally, Each motor provides a torque of 200 oz-in, estimate of the robot tilt angle from gyroscope and
sufficient for our purpose. The maximum operating current accelerometer measurements.
and voltage ratings are givenas 14 Ampere and 16 Volts.
C. Arduino Mega Development Board
Selection of the development board is based on
the following features:
1) Performance: Normally, the self-balancing robot needs
almost real-time response to estimate and correct its tilt
angle.Hence, the development of board must provide a
processing speed that is sufficiently fast to perform the
processing tasks, including data acquisition, control
computation and signal output, within the sampling time.The
Arduino Mega development board is equipped with the
ATmega328 processor, which features a maximum clock rate
of 16 MHz.
2) Input / Output Pins:On the robot, sensors are deployed to
obtain measurements of its motion: a gyroscope and an
accelerometer are used to estimate the tilt angle,encoders are
used to obtained the measurements of the robot. The Arduino Fig 3: Tilt velocity of the robot body

All Rights Reserved © 2018 IJEREEE 12

 ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

A. Angle Estimation via Complementary Filter

In Section II, a gyroscope is used to measure the angular
velocity of the robot. And accelerometer measures the Y - and
Z-components of gravitationalacceleration, encoders measure
the distancetravelled by the wheels.For thebalacning control,
the tilt angle is corrected to the upright position. Balancing
control is possible by using wheel motion of the robot. In
order to solve the problem of angle measurement, we can
design a complementary filter [13], a block diagram of which
is as shown in Fig. 3.

Fig. 6. Tilt angle estimation.

Here Blue line represents: output value of

complementaryfilter; red line represents: direct integration of
Fig. 4. Block Diagram of the Complementary Filter gyroscope angular rate; green line represents: angle deduced
from accelerometer measurement using the direction
Next tuning the filter, its performance under noise isfirst tested cosinemethod.
in simulation as shown in Fig. 4, whichcompares the three
ways to estimate the tilt angle. As we know, the best estimate B. Synchronization of a Inner-loop Control for Wheel
of the angle is provided by the complementary filter. Here, we If we can assume ideally, if the same motor control signal is
can see the same result is also observed in experiment, as sent to the motors control board, both motors should rotate at
shown in Fig.5 the same speed. A method is used to synchronize both
motors is needed.

Fig. 7. Block diagram of wheel synchronizer.

Fig 5: Sense tilt and drive wheels to make robot erect

Fig. 8. Test results of wheel travel distances.

All Rights Reserved © 2018 IJEREEE 13

 ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

This gain uses the derivative of the error to reduce the

Top: The wheels’travel distances coincide when wheel overshoot of the output erivative gain This gain uses the
synchronization control is employed. derivative of the error to reduce the overshoot of the output.
Bottom: At equal motor input, the wheels travel at different erivative gain This gain uses the derivative of the error to
speed withwithout wheel synchronization. reduce the overshoot of the output.

C. Balancing Control Designs Derivative gain:

Here In this project, two control design strategies are studied The gain is uses the derivative of the error to reduce the
one is proportional-integral-differential (PID) control, and overshoot of the output. The three gains, kP, kI and kD are
proportional-integral proportional-differential (PI-PD) control tunable parameters and must be adjusted according to the
[6] based on linear-quadratic-regulator (LQR) design. specific hardware used. There are very many ways to tune
these parameters to an optimal value. We used the manual
method to do this. First, we set all the gains to 0. Then kP is
increased until the response oscillates. Then kP is set to half
its value. kI is now increased until the steady state error is
corrected in sufficient time. Finally, kD is increased to
minimize overshoot.
2) LQR, Linear Quadratic Regulator

To control a linear system ,LQR-control can be applied. The

states of the dynamical system are described by a state space
model including the matrices A, B and C as seen in Figure 11:

Fig. 9. Definition of motion variables

1) PID Control:

Fig 11: Block representation of state space system showing

the system matrices A, B and C as well as the gain matrix K.

The Design of the Self-Balancing Robot LQR Controller. The

LQR method is the most mature controller design method in
the development of modern control theory [15]; LQR optimal
Fig. 10. Block diagram of PI-PD controller
control is to seek the control amount 𝑢∗(𝑡) to make the system
The block diagram of the PID controller is as shown in Fig.
reach the steady state and guarantee the performance index 𝐽
10, where the complementary filter generates estimates of the
to take the minimum value:
angular velocity and angle.
𝐽 = ∫ ∞ 0 (𝑋𝑇𝑄𝑋 + 𝑢𝑇𝑅𝑢) 𝑑t
Proportional gain:
This gain directly relates to the current error to the current The parameter of the robot is shown in below table:
motor output. This forms the bulk of the gain in the controller
and reacts instantly to any error
Integral gain:
This gain compensates for any steady state error caused by
drift or the offsets in the center of mass of the robot. It works
by integrating the values of error over a period of time and
hence reacts to steady state error after it accumulates.
Derivative gain

All Rights Reserved © 2018 IJEREEE 14

 ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

IV. EXPERIMENTAL RESULTS

The figure shown below shows the working of the Self

Balancing Robot with Ardunio Microcontroller

Fig 12: Combined outputs Series1-gravity angle(only

accelerometer reading) Series2-stabilized angles(combined
readings)

V. CONCLUSION

This paper presents an experimental, Arduino based, low cost

self-balancing robot as an educational control system.Low-
cost communicate hardware such Xbee may be added to the
robot, so that the user can control its motion remotely and read
the data via telemetry. Also, user-controlled motion such as
forward, backward, turn right and left, may be implemented.A
possible extension is to combine the Arduino system and other
sensors such ultrasonic and IR senses, GPS, digital compass
and Camera to address other advanced applications, e.g.
obstacle avoidance and perimeter following. The main
drawback that hampers the overall project result its unable to
attain. Therefore appropriate conclusions are not able to
achieve. The problem with the oscillation still remains with
the system and future work has to be done to achieve a stable
solution.

VI. FUTURE IMPROVEMENT

The stabilization provided by the reaction wheel is limited be

the torque provided by the reaction wheel motor. Subsequent
plan is to use a rotating disc and its gyroscopic precession for
balancing. This would provide a more stable design capable of
providing higher restoring torque .In such a case particular
attention should be paid to any rotary axes, their alignment,
and how they are fixed to the model, to the position and
alignment of brackets, and to the mounting and fastening of
any flexible couplings. In addition to this, fuzzy logic
Fig 11: Individual outputs of accelerometer and gyro controller can also be implemented to provide flexibility and
accuracy in control.

All Rights Reserved © 2018 IJEREEE 15

 ISSN (Online) 2395-2717

International Journal of Engineering Research in Electrical and Electronic

Engineering (IJEREEE)
Vol 4, Issue 2, February 2018

REFERENCES [12] V. Georgitzikis and I. Akribopoulos,

O.andChatzigiannakis, “Controlling physical objects via the
[1] Arduino. [Online]. Available: http://arduino.cc internet using the Arduino platform over 802.15.4 networks,”
IEEE Latin America Transactions, vol. 10, no. 3, pp. 1686–9,
[2] D. Arndt, J. E. Bobrow, S. Peters, K. Iagnemma, and S. 2012.
Bubowsky, “Two-wheel self-balancing of a four-wheeled
vehicle,” IEEE Control Systems Magazine, vol. 31, no. 2, pp. [13] A.-J. Baerveldt and R. Klang, “A low-cost and low-
29–37, April 2011. weight attitude estimation system for an autonomous
helicopter,” in IEEE International Conference on Intelligent
[3] R. Fierro, F. Lewis, and A. Lowe, “Hybrid control for a Engineering Systems, sep 1997, pp. 391–5.
class of underactuated mechanical systems,” IEEE
Transactions on Systems, Man and Cybernetics, Part A: [14] R. C. Ooi, “Balancing a two-wheeled autonomous robot,”
Systems and Humans, vol. 29, no. 6, pp. 649–4, nov 1999. Final Year Thesis, The University of Western Australia,
School of Mechanical Engineering, 2003.
[4] K. Xu and X.-D.Duan, “Comparative study of control
methods of single-rotational inverted pendulum,” in [15] L. Cheng, H. Dewen, P. Yaodong, Z. Xiaocai, and D.
Proceedings of the First International Conference on Machine Guohua, “Fuzzy control of a quintuple inverted pendulum
Learning and Cybernetics, vol. 2, 2002, pp. 776–8. with the LQR method and 2-ary fuzzy piecewise interpolation
function,” in Proceedings of the 45th IEEE Conference on
[5] F. Grasser, A. D’Arrrigo, S. Colombi, and A. C. Rufer, Decision and Control (CDC ’06), pp. 6307–6312, December
“JOE: A mobile, inverted pendulum,” IEEE Transactions on 2006.
Industrial Electronics, vol. 49, no. 1, pp. 107–14, 2002.
[6] D. P. Anderson. (2003, Aug.)nBot balancing robot.
Online.[Online]. Available: http://www.geology.smu.edu/
dpa-www/robo/nbot

[7] J. Sarik and I. Kymissis, “Lab kits using the Arduino

prototyping platform,” in IEEE Frontiers in Education
Conference, 2010, pp. T3C– 1–5.

[8] G. Guo and W. Yue, “Autonomous platoon control

allowing rangelimited sensors,” IEEE Trans. on Vehicular
Technology, vol. 61, no. 7, pp. 2901–2912, 2012.

[9] P. A. Vignesh and G. Vignesh, “Relocating vehicles to

avoid traffic collision through wireless sensor networks,” in
4th International Conference on Computational Intelligence,
Communication Systems and Networks. IEEE, 2012.

[10] M. J. A. Arizaga, J. de la Calleja, R. Hernandez, and A.

Benitez, “Automatic control for laboratory sterilization
process rsd on Arduino hardware,” in 22nd International
Conference on Electrical Communications and Computers,
130-3, Ed., 2012.

[11] S. Krivic, M. Hujdur, A. Mrzic, and S. Konjicija, “Design

and implementation of fuzzy controller on embedded
computer for water level control,” in MIPRO, Opatija,
Croatia, May 21–25 2012, pp. 1747–51.

All Rights Reserved © 2018 IJEREEE 16