Fuzzy PID for Quadrotor Space Fixed-Point
Position Control
Jie MA
Harbin Institute of Technology
Control and Simulation Center
Harbin, Heilongjiang province
Email: majie@hit.edu.cn
Ruihang JI
Harbin Institute of Technology
Control and Simulation Center
Harbin, Heilongjiang province
Email: jiruihang@126.com
Abstract—
Quadrotor is widely used in
the military or civil,
which thanks to its high maneuverability, low maintenance cost
and easy research. In real life, space fixed-point position control
is necessary and essential because of the outdoor GPS or indoor
visual positioning. All kinds of research and development are
based on the great control performance. However, under
actuated, strong coupling and nonlinear properties makes us
cannot get accurate models of quadrotor. How to get a smooth
and efficient performance for the quadrotor space fixed-point
position control in a variety of conditions attracts a lot of
researchers. According to the features of quadrotor, this paper
we put forward fuzzy PID controller for space fixed-point
position control. This fuzzy PID controller is widely used in cases
where mathematical models of system cannot be get in accurate
modeling with nonlinear. Based on fuzzy-set theory, fuzzy
language variables and fuzzy logic-based intelligent computer
control, this make the fuzzy PID have the ability of self-tuning
environment. By comparing the
traditional PID and fuzzy PID for quadrotor space fixed-point
position control with Matlab simulation and actual flight test,
fuzzy PID perform well than traditional PID.
Keywords—Fuzzy PID; Fixed-point Control; quadrotor;
Mathematic Model
I. INTRODUCTION
A quadrotor is a cross-shaped unmanned aerial vehicle
which can achieve six degrees of freedom movement by
changing
its four engines’ speed. This construction makes the
quadrotor can take off vertical and perform well in a narrow
space. So high maneuverability, simplicity of construction,
low maintenance costs and low noise are the advantages of
this unique aerial vehicle[1], which makes the quadrotor has
broad prospects in aerial surveying and mapping, military
reconnaissance. Space fixed-point position control for
quadrotor is necessary and essential with the outdoor GPS and
indoor visual positioning. How to perform smooth and
efficient for this control process attract a lot of researchers’
attention.
4 Gao Qing used LQR optimal control which gave
good results at lower speeds, but if there are a large
maneuverable this method provided a poor performance. But
turbulence and uncertainty are not taking into account, so this
method cannot meet desired control in some cases[2].
Li Jun used the PID controller to control quadrotor[3],
laboratory researchers compared the results with the LQR
controller. From the result, PID controller perform well for
quadrotor hovering state but if strong turbulence is added to
the system, this controller show poor performance.
In these studies, PID and LQR controller are designed on a
linear and actual dynamic model. But in reality, most of
systems accompanied by nonlinear, time-varying uncertainty
make us cannot get the actual dynamic model.
So Fuzzy PID controller for quadrotor is put forward to the
quadrotor, because of the features of quadrotor. All the drags,
aerodynamic, Coriolis and gyroscopic effect are neglected[4].
So from the dynamic model of quadrotor which we neglected
a number of factors, under actuated, strong coupling and
nonlinear properties obvious[5-6], Fuzzy PID with three
parameters which according to the current statement of system
alter value by self-tuning. To some extent, this controller has
the ability of self-learning[7-8], although nonlinear, strong
coupling systems.
This paper we will build a dynamic model of quadrotor, to
recognize where the features of quadrotor come from. Then
comparing the traditional PID and Fuzzy PID controller for
quadrotor space fixed-point position control from the
simulation data. Finally using the fuzzy PID theory and
traditional PID on our quadrotor platform in practice, shows
the feasibility of this theory.
II. MATHEMATIC MODEL
In order to obtain the dynamic modeling of quadrotor, we
establish two coordinates system: inertial
frame and body frame in Figure 1.
x
b
Yb
Zb
x
E
Y
E
Z
E
Figure 1 The quadrotor schematic
2016 Sixth International Conference on Instrumentation & Measurement, Computer, Communication and Control
978-1-5090-1195-7/16 $31.00 © 2016 IEEE
DOI 10.1109/IMCCC.2016.131
721