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

net/publication/281619233

The quadrotor MAV system using PID control

Conference Paper · August 2015

DOI: 10.1109/ICMA.2015.7237537

CITATIONS READS
34 1,050

3 authors, including:

Minh Nguyen Duc Thang Nguyen Trong

Thuy Loi University Thuy Loi University
10 PUBLICATIONS 87 CITATIONS 37 PUBLICATIONS 274 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Thang Nguyen Trong on 09 September 2015.

The user has requested enhancement of the downloaded file.

 Proceedings of 2015 IEEE
International Conference on Mechatronics and Automation
August 2 - 5, Beijing, China

The Quadrotor MAV System using PID Control

Minh Nguyen Duc Thang Nguyen Trong Yang Sheng Xuan
Department of Electrical & Electronics Department of Electrical & Electronics Electronic Information and Electrical
Haiphong Private University Haiphong Private University Engineering Faculty
Haiphong, Vietnam Haiphong, Vietnam Dalian University of Technology
[email protected] [email protected] Dalian,China

Abstract: Quadrotor MAV is a simple structure multi-rotor research in it, and many countries are placing the military
aircraft with vertical take-off and landing ability, rendering it applications of UAV as a top priority.
suitable for ground environmental inspection in both civil and For fixed-wing UAV also known as, VTOL (Vertical
military domain. This paper focuses on mathematical modeling Take-Off and Landing) UAV, rotorcraft development is much
and flight controller design of the quadrotor.
According to the Newton-Euler formalism, the generic 6 DOF
slower. This is because the VTOL UAV control is more
dynamic motion equation is given for control algorithms and complex, the level of early technology cannot achieve its
further research. In order to achieve the goal of flight control, autonomous flight control. However, fixed-wing VTOL UAV
there are two kinds of control systems based on the PID has significant advantages [3][7][8] such as: possesing
algorithm. Due to the under actual characteristic, the independent takeoff and landing capability for adaptation with
approximate linear model is given by adding the assumed the environment, having ability to fly in a variety of modes,
condition. Because of the simplified model, the single-loop PID namely hover, forward flight, flying side and back flying and
control system has been designed, in order to keep the attitude so on. These advantages have made the fixed-wing VTOL
stabilization of quadrotor in the target height. In context of UAV suitable for broader range of applications. In the civil
position control, the double-loop PID control system has been
designed for the achievement of these objectives - the position of
context, VTOL UAV has broad application prospects in traffic
quadrotor easily made to reach the target position accurately and monitoring, resource exploration, power line inspection, forest
the stabilization of quadrotor is maintained when hovering. The fire prevention and other aspects.
whole double-loop control system is separated into two parts- the In VTOL UAV, the Quadrotor aircraft with its new
inner loop control (attitude control) and the outer loop control structure and layout [10], and having a unique way of flying
(position control). has raised widespread concern among the researchers.
There are two kinds of control systems as discussed in this Compared with conventional single-rotor, Quadrotor machine
research, and these systems have been tested in a simulation structure is more simple, its propeller has a fixed ange of
environment. The validity of the control system design is verified attack, and for the propeller, only changing the rotational
by analyzing simulation results. Attitude control experiment is
implemented in the experiment platform. The results of
speed of the four rotor can change the lifting force , thereby
experiment verify that the PID control strategy can achieve the changing the attitude and position of the aircraft to achieve a
stabilization control of the quadrotor. variety flight control [2][5]. The payload for quadrotor is
higher than the single-rotor as it can produce more lift and
Index Terms- Quadrotor, Robot control, PID control. thus have greater load capacity. Its symmetric structure, due to
the presence of the four rotor, can be of help for cancellation
I. INTRODUCTION of trans twisting moment produced by each rotor and so
Along with the rapid development of satellite and quadrotor does not require special anti-torque propeller [16],
Unmanned Aerial Vehicles (UAV) technology, MEMS thus leading to the reduction of the aircraft's power
technology has been widely used in all aspects of defence, consumption. So, quadrotor has a improved energy efficiency.
military and civilian products, processes and information Because of its wonderful structure, it can be used for making
technology. The miniaturization of aircraft has caused the lower propeller speed quadrotor aircraft, with even in-
dramatic rise in the extensive use of unmanned aerial vehicles flight access to the blade barrier. So, for this condition, it will
(UAV), and this has been a relevant topic of research in many not cause as much destruction as the single-rotor aircraft,
countries. UAV refers to a power unit which operates without rendering it suitable for a relatively narrow space flight with
the aircraft personnel inside it. Some physical aspects good mobility.
associated with it are the use of air power to overcome its own These advantages make the quadrotor aircraft inherit broad
weight, remote control or autonomous flight, capability of prospects for military and civilian applications, especially for
carrying the load being either one-time or multiple recycling. reconnaissance in near-surface environments (such as indoor,
UAV possesses the features such as a high-quality, near real- jungle and urban, etc.), monitoring and other tasks. In addition,
time, all-weather reconnaissance, surveillance, target the new appearance, low cost, simple structure, excellent
acquisition, interception and battle damage assessment performance and a unique way to make the majority of the
missions, or even directly attack important targets. Given the flight control has made it an important topic of study among
widespread use of UAV and outstanding performance in the the researchers.
war, the international community has set off a new wave of

978-1-4799-7096-4/15/$31.00 ©2015 IEEE 506

 II. BACKGROUND QUADROTOR DYNAMIC MODEL c c c s s  s c c s c  s s 
A. Quadrotor model. R  c s s s s  c c s s c  c s  (1)
 s c s c c 
A quadrotor is an under actuated aircraft with fixed pitch
angle four rotors as shown in Figure 1.
In the quadrotor, there are four rotors with fixed angles where Sθ = sin(θ), Cψ = cos(ψ), etc., and R is the matrix
which represent four input forces that are basically the thrust transformation.
generated by each propeller as shown in Figure 1. Using the notation of [12-16] we introduced a quadrotor
The collective input is the sum of the thrusts of each motor. dynamic model as follow:
Pitch movement is obtained by increasing (reducing) the speed
of the rear motor while reducing (increasing) the speed of the  x  (sin  sin   cos sin  cos  ) U1
front motor. The Roll movement is obtained similarly by  m
increasing (reducing) the speed of the right motor while  U1

 y  (  cos sin   sin  sin  cos  ) m
reducing (increasing) the speed of the left motor. The Yaw
movement is obtained by increasing (decreasing) the speed of  U1
the front and rear motors together while decreasing z   g  (cos  cos  )
 m (2)
(increasing) the speed of the lateral motors together [19].  I  Iz JTP U2
p  y qr  q 
 I Ix Ix
 I x I JTP U3
q  z x
pr  p 
 Iy Iy Iy
r  I x  I y pq  U 4
 Iz Iz

where m is the helicopter mass, g is the gravitational

acceleration, JTP is moment of inertia of the rotor:
U 1  b ( 12   22   32   24 )
 2 2
Fig.1 Quadrotor Coordinate System. U 2  bl ( 4   2 )
 2 2 (3)
where (x, y, z) denote the position of the center of mass of the U 3  bl ( 3   1 )
rotorcraft relative to the frame and the roll, pitch and yaw  2 2 2 2
U 4  dl ( 2   4   1   3 )
Euler angles which represent the orientation of the craft are     1   2   3   4

noted as ϕ, θ and ψ, respectively, and 1 ,  2 , 3 ,  4 denote a
motor speed signal. III. THE DESIGN OF PID CONTROLLER FOR QUAROTOR
B. Newton-Euler equations. Laplace transform traditional PID structure as follow:
Rigid motion can be divided into transitional and rotational. Ki
u (s)  ( K p   sK d )e( s ) (4)
Transitional equation : s
Set r is centroid position, m for rigid quality, according to where e denoted error of input and output, u denoted control
d signal, Kp, Ki, Kd denoted coeficient of gain, time integral,
Newton's second law, we have f  ( mr)  mr time derivative, respectively.
dt
The controller introduced in figure 2 consist of 3 parts
d
Movement equation:  e  ( I  ) where  e denoted
dt
coordinate system is applied to the rigid torque, ω denoted
angular velocity of a rigid and I denoted the unit matrix.
C. Quadrotor dynamic model.
Quadrotor properties have a nonlinear, multivariable,
highly coupled, due to the drive system (6 degrees of freedom Fig.2 Block diagrams of the singer-loop control structure
and 4 inputs). In order to move the quadrotor model from the Part 1: Control algorithm is the main of the controller and
earth to a fixed point in the space, the mathematical design the input signal of the PID controller is error of Setpoint
should depend on the direction cosine matrix as in equation 1: module and output of the Sensor module, this part will be
described carefully in the next works.

507
 Part 2: Inverse transformation module: signal  ,  , and passed to the IMU module. The IMU
A simple dynamic model describe as: module is an inertial measurement unit to detect height signal
z and attitude signals  ,  , and feedback to the controller.
 z   g  (cos  cos  )U1 / m
  U 2 / I x
  U / I (5)
  U3 / Iy
 4 z

After inversed we have model follow as:

U1  ( z  g ) m / (cos  cos  )

U 2  I x
U 3  I y (6)
U 4  I z
Part 3: By formula (3) can be obtained Ω2 from U by inverse
transform:
Fig.4 Control module internal structure – Controller
2  1 U  1 U  1 U
 1 4b 1 2bl 3 4d 4 In the Figure 4, In1 is the setpoint of the height (z) of the
 2 1 1 1 system, In2 is the height actual value which measured from
2  4b U1  2bl U 2  4d U 4 the system. The error of In1 and In2 is the input signal of z
 (7) PID controller. In3 is the set attitude signal value of the
1 1 1
32  U1  U3  U4 system, In4 is the actual attitude signal value which measured
 4b 2bl 4d from the system. The error of In3 and In4 is the input signal of
24  1 U1  1 U 2  1 U 4 ϕ PID controller, θ PID controller and ψ PID controller. The
 4b 2bl 4d 4 signals PID controllers are according to equation (6),
through the inverse transform to obtain the value of the height
IV. SIMULATION RESULTS control and attitude control and transform into control signals
U1, U2, U3 and U4. And then according to equation (7) we
A. Simulation System.
inverse transform that control signals to obtain
In this work, the way to build the PID controller for
quadrotor is presented and selected control software is the 1 ,  2 , 3 and  4 .
xPC Target environment with MATLAB/Simulink.

Fig.3 PID simulation system

In the Figure 3, the input module is the input of system, the

control module is shown in the Figure 4, it is based on the
given values and measured values constitute a control
deviation, for PID control deviation, by the inverse transform
and highly rotor speed control will be controlled and attitude
control output value is converted into a motor speed signal Fig.5 Module internal structure - quadrotor dynamics model.

1 ,  2 , 3 ,  4 . The motor speed signal is provided by the In the Figure 5, the input Omega_i module is the actuator
controller in the motors module. The selector module with motor speed signal is given and according to the formula (3),
input from the quadrotor aircraft dynamic model (shown in through the Omega to U to converts U1, U2, U3 and U4. Then,
Figure 5), it is selector to select the height signal z and attitude according to equation (2) to build quadrotor aircraft dynamics
model.

508
B. The simulation results and analysis.
In this work, simulation results are presented wherein
several step signals are applied as a reference for the triplet of
angles, as well as for the height.
We setting the initial state are ϕ = 0.2rad, θ = 0.2rad, ψ =
0.5rad. The sensor noise does not consider the impact of the
model in Simulink performance simulation when testing
single-loop control system. After several rounds of testing, the
control parameters for each channel values as shown in Table
1.
TABLE 1
Single-loop control system simulation parameter value table
Kp Ki Kd time (10-2 s)
Fig.8 Stability control system response of Yaw.
Roll 0.75 0 0.15
Pitch 0.75 0 0.15
Yaw 0.4 0.02 0.2
Hight 23.5 6.0 19.5

Running the system, the results shown in figures:

time (10-2 s)
Fig.9 Stability control system response of z.

As can be seen from the simulation results, in this initial

conditions, the quadrotor aircraft eventually stabilize at the set
height z = 0.5m, three attitude angles after shock adjustment to
stabilize within 3s, reaching attitude angle are zero
requirements, and basic no steady-state error. Thus the
time (10-2 s)
Fig.6 Stability control system response of Roll
controller designed using this method is effective. But the
dynamic characteristics of the roll channel and pitch channel
are not very satisfactory, larger overshoot, oscillation cycle
more, whereas the control yaw channel is better.
V. CONCLUSION
In this paper a PID controller is designed by singer-loop
control structure for the pitch, roll and yaw respectively for
the quadrotor helicopter has been presented. And the
simulation results show that the controllers have good
adaptive ability and robustness. But single-loop system has
only 4 control signals, so it only controls the hight and
stability. If in practical problems, we need to know carefully
the orbit of quadrotor to achieve it can flight from point to
point. That is the quadrotor exact reaches the setpoint, and
then remain in stability state. If want to achieve this, we need
time (10-2 s) to control six variables (x, y, z, ϕ, θ, ψ), but the single-loop
Fig.7 Stability control system response of Pitch PID control system can not be achieved. Therefore, we need to
double-loop PID control system to do that. And the double-

509
loop PID control system will be introduced in part II of the [10] Salvador Gonz´alez-V´azquez and Javier Moreno-Valenzuela, “A
New Nonlinear PI-PID Controller for Quadrotor Posture Regulation”,
next paper. Electronics, Robotics and Automotive Mechanics Conference, pp
REFERENCES 642-647, 2010.
[11] Fatih Şenkul and Erdinç Altuğ, “Adaptive control of a tilt - roll rotor
[1] Sa A.Al-Hiddabi, “Quadrotor Control using Feedback linearization quadrotor UAV”, International Conference on Unmanned Aircraft
with Dynamic Extension”, Proceeding ofthe 6th International Systems (ICUAS), pp 1132-1137, 2014
Symposium on Mechatronics and its Applications (ISMA09), pp 01- [12] Swee King Phang, Chenxiao Cai, Ben M. Chen, Tong Heng Lee,
03, 2009. “Design and mathematical modeling of a 4-standard-propeller (4SP)
[2] Gabriel M.Hoffmann, Haomiao Huang, Steven L.Waslander, Claire J. quadrotor”, Proceedings of the 10th World Congress on Intelligent
Tomlin, “Quadrotor helicopter flight dynamics and contro-Theory Control and Automation, pp 3270-3275, 2012.
and experiment”, AIAA Guidance, Navigation and Control [13] Shahida Khatoon, Mohammad Shahid, Ibraheem, Himanshu
Conference and Exhibit, pp 01-20, 2007 Chaudhary, “Dynamic modeling and stabilization of quadrotor using
[3] Abdelhamid Tayebi and Stephen McGilvray, “Attitude Stabilization PID controller”, International Conference on Advances in Computing,
of a VTOL Quadrotor Aircraft”, Ieee Transactions On Control Communications and Informatics (ICACCI), pp 746-750, 2014.
Systems Technology, Vol. 14, No. 3, pp 562-571, 2006. [14] A Hyunsoo Yang and Dongjun Lee, “Dynamics and control of
[4] Erdinç Altu˘, James P. Ostrowski, Camillo J. Taylor, “Control of a quadrotor with robotic manipulator”, IEEE International Conference
quadrotor helicopter using dual camera visual feedback”, The on Robotics & Automation (ICRA), pp 5544-5549, 2014.
Iernational Jounal Of Rbotics Reseach, pp 329-341, 2005. [15] Fatih Senkul and Erdinç Altuğ, “Modeling and Control of a Novel
[5] A. Benallegue1, A. Mokhtari and L. Fridman, “High-order sliding- Tilt – Roll Rotor Quadrotor UAV”, International Conference on
mode observer for a quadrotor UAV”, International Jounal of Robust Unmanned Aircraft Systems (ICUAS), pp. 1071–1076, 2013.
and Nonlinear Control, pp 427-440, 2008. [16] Markus Ryll, Heinrich H. B¨ulthoff, and Paolo Robuffo Giordano,
[6] Odile Bourquardez, Robert Mahony, Nicolas Guenard, Franc¸ois “Modeling and control of a quadrotor UAV with tilting propellers”,
Chaumette, Tarek Hamel, and Laurent Eck, “Image-Based Visual IEEE International Conference on Robotics and Automation, pp
Servo Control of the Translation Kinematics of a Quadrotor Aerial 4606-4613, 2012.
Vehicle”, IEEE Transactions On Robotics, Vol. 25, No. 3, pp. 743– [17] Norafizah Abas, Ari Legowo, Rini Akmeliawati, “Parameter
749, 2009. identification of an autonomous quadrotor”, 4th International
[7] Travis Dierks, and Sarangapani Jagannathan, “Output Feedback Conference on Mechatronics (ICOM), 2011.
Control of a Quadrotor UAV Using Neural Networks”, IEEE [18] Nguyen Dang Hao, Boutayeb Mohamed, Hugues Rafaralahy,
Transactions On Neural Networks, Vol. 21, No. 1,, pp. 50–66, 2010. “Trajectory-tracking control design for an under-actuated quadrotor”,
[8] C. Nicol, C.J.B. Macnab, A. Ramirez-Serrano. “Robust adaptive European Control Conference (ECC), pp 1765-1770, 2014.
control of a quadrotor helicopter”, C. Nicol et al. / Mechatronics 21, [19] Atheer L. Salih, M. Moghavvemi, Haider A. F. Mohamed and Khalaf
pp. 927–938, 2011. Sallom Gaeid, “Flight PID controller design for a UAV quadrotor”,
[9] Justin M.Selfridge and Gang Tao, “A multivariable adaptive Scientific Research and Essays Vol. 5(23), pp. 3660-3667, 2010.
controller for a quadrotor with guaranteed matching conditions”,
American Control Conference (ACC), pp 26-31, 2014.

510

View publication stats