Design and Implementation of Self-tuning Control
Method for the Underwater Spherical Robot
Abstract-Considering the complicated disturbance in
underwater circumstance, usually it is difficult to solve the
control problem when the robot changes its motion state or it is
subject to ocean currents, its performance deteriorates since the
fixed set of parameters is no longer valid for the new conditions.
Thus, in this paper, an auto-tune PID (Proportional + Integral +
Derivative)-like controller based on Neural Networks is applied
to our amphibious spherical underwater robot, which has a great
advantage on processing online for the robot due to their
nonlinear dynamics. The Neural Networks (NN) plays the role of
automatically estimating the suitable set of PID gains that
achieves stability of the system. The NN adjusts online the
controller gains that attain the smaller position tracking error.
The performance of the NN-based controller is investigated in
ADAMS and MATLAB cooperative simulation. The velocity of
the spherical robot can be controlled to precisely track desired
trajectory in body-fixed coordinate system. Additionally, real
time experiments on our underwater spherical robot are
conducted to show the effectiveness of the algorithm.
Index Terms - Underwater Spherical Robot; Virtual
Prototype; Neutral Network PID Control; Auto-tuning;
Cooperative Simulation
I. INTRODUCTION
Underwater robots have been widely used in many
subsea tasks, ranging from ocean inspection to repair of
underwater structures related mainly to the power and oil
industry. Very often, according to the task, the underwater
robot is required to continuously change its operating tool or
to pick up and release loads causing a change in behaviour [1]-
[10]. That results as an inherent change in its weight,
buoyancy and hydrodynamic forces; and as a consequence, a
decrease in the position tracking performance. In addition,
underwater robots have to deal with the highly dynamical
underwater environment represented in the form of ocean
currents and waves in shallow water. With this in mind, when
the dynamic characteristics of the system are time dependent
or the operating conditions of the system vary, it is necessary
to re-tune the gains to obtain the desired performance,
resulting in time consumption.
As a traditional method, conventional PID controller is
simple and practical, but has the disadvantage of the difficulty
of adjusting parameter online [11]. NN control theory has
been widely used in underwater control system. A 3-layer NN
controller has been constructed for underwater robot by Yuh
[12], and based on this, Lorenz [13] does research upon an
underwater robot of moving up and down. Neutral network
control has the strong ability of information integration and
complex system control. But neutral network also has some
weakness which restricts its development. For instance, it has
a slow convergence rate and a long training time, which
couldn’t be accepted by most control system. Also, traditional
neutral network doesn’t satisfy the performance index of a
control system, including fast response, less overshoot and so
on.
Consequently, a variety of motion control methods have
been proposed and the intelligent control techniques include
PID control, fuzzy control, adaptive control, neural network
control, or a mix of them. Research [14]-[16] present systems
with a mix of neural networks and fuzzy control in which the
training and rules of behaviour are based on the desired states.
Their performance is described as accurate when uncertainty
and perturbations take place while performing a trajectory.
Although the training periods are extremely long, there are
also combinations of PID controls and a smart system aimed
to auto-tune the gains of different systems such as [17]-[22].
In this paper, an auto-tune PID-like controller based on
an online NN is implemented on our spherical underwater
robot; for trajectory tracking with unknown disturbances.
Simulation results are given considering the non-linear
hydrodynamics of robot; including disturbances of ocean
currents. Real time experiments on spherical underwater robot
are conducted to show the effectiveness of the proposed
scheme. For the remaining sections of this paper in Section 2
the general system model of underwater robot and the effect of
ocean currents are presented, Section 3 presents the neural
network-based self-tuning PID control, Section 4 describes the
co-simulation based on ADAMS and MATLAB and
experimental results; Finally in Section 5 the conclusions and
future works are provided.
II.
MODELING OF THE SPHERICAL UNDERWATER ROBOT
As introduced in references [1]-[6], we developed an
amphibious spherical robot capable of motion on land and
underwater to perform complicated operations. The
underwater movement mechanisms of this amphibious
spherical robot are the same as those presented in previous
Yanlin He
1, 2
,
Shuxiang Guo
1, 2, 3*
,
Liwei Shi
1, 2 *
,
Huiming Xing
1
, Zhan Chen
1
, Shuxiang Su
1
1
Key Laboratory of Convergence Medical Engineering System and Healthcare Technology,
the Ministry of Industry and Information Technology, Beijing Institute of Technology,
No.5, Zhongguancun South Street, Haidian District, Beijing 100081
2
Key Laboratory of Biomimetic Robots and Systems, Ministry of Education, Beijing Institute of Technology, No.5,
Zhongguancun South Street, Haidian District, 100081 Beijing, China.
3
Faculty of Engineering, Kagawa University, 2217-20 Hayashi-cho, Takamatsu, Kagawa 760-8521, Japan
Email: heyanlin@bit.edu.cn, guoshuxiang@bit.edu.cn, shiliwei@bit.edu.cn
* Corresponding author
978-1-5090-6759-6/17/$31.00 ©2017 IEEE
Proceedings of 2017 IEEE
International Conference on Mechatronics and Automation
August 6 - 9, Takamatsu, Japan