sensors

Article Thiết kế và phát triển

Design and Development of an Autonomous Underwater
Helicopter for Ecological Observation of Coral Reefs
Jing Zhou 1,2,3 , Nanxi Zhou 2,3 , Yuchao Che 3 , Jian Gao 2 , Liming Zhao 3 , Haocai Huang 1,4 and Ying Chen 1,4, *

1 Hainan Institute, Zhejiang University, Sanya 570203, China; jingzhou@zju.edu.cn (J.Z.);

hchuang@zju.edu.cn (H.H.)
2 Polytechnic Institute, Zhejiang University, Hangzhou 310015, China; 22160113@zju.edu.cn (N.Z.);
gjsy1202@163.com (J.G.)
3 College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China; cyc521521@zju.edu.cn (Y.C.);
0921201@zju.edu.cn (L.Z.)
4 Ocean College, Zhejiang University, Zhoushan 316021, China
* Correspondence: ychen@zju.edu.cn

Abstract: Real-time status monitoring is an important prerequisite for coral reef ecological protection.
kiến trúc hệ thống, thiết bị điện tử,
cảm biến và bộ truyền động Existing equipment does not provide an ocean observation platform with adequate mobility and
-> Giải thích thuật toán đường dẫn và efficiency. This paper describes the design considerations of a proposed autonomous underwater
bộ điều khiển helicopter (AUH) dedicated for ecological observation of coral reefs, including the system architecture,
electronic devices, sensors and actuators, and explains the path control algorithm and controller to
follow a specific path for ocean exploration. The structure and dynamic model of the AUH are first
introduced, and then the corresponding simplification is made for motion analysis. Furthermore,
computational fluid dynamics (CFD) simulation is carried out to evaluate the dynamic performance





 of the AUH. Fuzzy-PID control algorithm is utilized to achieve a good antidisturbance effect. In order
Citation: Zhou, J.; Zhou, N.; Che, Y.; to validate the performance of the proposed underwater vehicle, a field test was performed, and
Gao, J.; Zhao, L.; Huang, H.; Chen, Y. results confirmed the feasibility of the proposed prototype.
Design and Development of an
Autonomous Underwater Helicopter Keywords: autonomous underwater vehicle; disk-shaped; sensors; dynamic performance; Fuzzy-PID
for Ecological Observation of Coral
Reefs. Sensors 2022, 22, 1770. https://
doi.org/10.3390/s22051770

Academic Editors: Amirmehdi 1. Introduction

Yazdani, Zheng Zeng,
Somaiyeh MahmoudZadeh
and Yuanchang Liu
Received: 30 January 2022
Accepted: 22 February 2022
Published: 24 February 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Coral reefs, known as “underwater rainforests”, are the most biodiverse ecosystems in
the world and provide considerable economic, ecological and social benefits for human
beings. With the increasing intensity of social and marine economic activities in coastal
areas, coral reefs are facing the threat of severe damage [1]. Real-time observation is an
important prerequisite for the rescue and protection of coral reefs; it helps to understand the
status of coral reef systems, identify relative impact factors and evaluate the effectiveness
of conservation measures, so as to put forward suggestions for further improvement on
protection strategies. Traditional coral observations are either carried out by divers carrying
cameras and samplers at the coral reef site or by fixed observation platforms with scientific
sensors [2,3], which have significant limitations in terms of observation scale and efficiency.
With the development of ocean technology, underwater vehicles equipped with professional
observation equipment are coming into application in the ocean observation field [4–11].
However, existing underwater vehicles do not provide an ideal platform for large-scale and
long-term observation of coral reefs. More specifically, (i) while an autonomous underwater
vehicle (AUV) mainly works in the upper water column, the observation of coral reefs
requires the underwater vehicle to be able to reach the seafloor and hover at a fixed point;
(ii) operation in the complex water flow near the seabed puts forward higher requirements
for the maneuverability of the underwater vehicle.
In view of the problems mentioned above, a seafloor-resident mobile observation
platform, Coral-AUH, is designed and developed in this paper. As a disk-shaped under-
water vehicle, the AUH has experienced several generations of iteration and come into

Sensors 2022, 22, 1770. https://doi.org/10.3390/s22051770 https://www.mdpi.com/journal/sensors

 underwater vehicle.
In view of the problems mentioned above, a seafloor-resident mobile observation
platform, Coral-AUH, is designed and developed in this paper. As a disk-shaped un-
Sensors 2022, 22, 1770 2 of 15
derwater vehicle, the AUH has experienced several generations of iteration and come in-
to application in various fields [12–18]. Coral-AUH inherits the super maneuverability
and seafloor-resident capability; meanwhile, it carries a camera and multiple sensors to
application in various fields
detect environmental [12–18].
data such Coral-AUH
as pH, inherits
conductivity, the superoxygen,
dissolved maneuverability
turbidity and
and
seafloor-resident
chlorophyll in real time, which helps scientists to determine the health status to
capability; meanwhile, it carries a camera and multiple sensors of detect
corals
environmental
more accurately. data
The such as pH,
advent of conductivity,
Coral-AUH will dissolved oxygen,
provide turbidity
a large-scale andand chlorophyll
long-term ob-
in real time, which helps scientists to determine the
servation platform for coral reef ecological observation. health status of corals more accurately.
The advent of Coral-AUH will provide a large-scale and long-term observation platform
The rest of this paper is arranged as follows: The general configuration of Coral-
for coral reef ecological observation.
AUH is discussed first, and then a kinematic model and motion analysis are presented.
The rest of this paper is arranged as follows: The general configuration of Coral-AUH
After that, computational fluid analysis evaluating the hydrodynamic performance of
is discussed first, and then a kinematic model and motion analysis are presented. After that,
Coral-AUH is presented. The control system is then described, and the results of a pool
computational fluid analysis evaluating the hydrodynamic performance of Coral-AUH is
test validating the feasibility of the control algorithm are presented. Finally, the results
presented. The control system is then described, and the results of a pool test validating the
of a field test verifying the observation capability and reliability of Coral-AUH are pre-
feasibility of the control algorithm are presented. Finally, the results of a field test verifying
sented.
the observation capability and reliability of Coral-AUH are presented.
Cấu hình
2. Configuration
2. Configuration of of Coral-AUH
Coral-AUH
2.1. System Architecture
the bottom-dwelling
Inspired by the bottom-dwelling stingray (Figure 1), we designeddesigned the
the disk-shaped
disk-shaped
autonomous underwater helicopter (AUH), (AUH), a new member of the AUV family, which is
seafloor-resident observation
suitable for seafloor-resident observation andand operation.
operation. Over
Over the decades, a series of
AUH prototypes
AUH prototypeshave
havebeen
beendeveloped
developed forfor various
various scenarios,
scenarios, andand in paper,
in this this paper, Coral-
Coral-AUH
is
AUHdesigned and implemented
is designed as the underwater
and implemented mobilemobile
as the underwater platform for coral
platform forobservations.
coral obser-
Coral-AUH
vations. Coral-AUH 0.5 m × 0.5
measuresmeasures 0.5mm×× 0.32
0.5 mm,× weighs
0.32 m,12 kg, is 12
weighs neutrally slightly positive-
kg, is neutrally slightly
buoyant, and can swim
positive-buoyant, for 60
and can minfor
swim at a60time.
min The
at a designed
time. Themaximum speed is 2 kn
designed maximum and is
speed the2
maximum
kn and thediving
maximum depth is 200depth
diving m. is 200 m.

Figure 1. The shape of the AUH inspired

inspired by
by bottom-dwelling
bottom-dwelling stingray.
stingray.

The
The system
system overview
overview of of Coral-AUH
Coral-AUH is is shown
shown in in Figure
Figure 2.2. It
It mainly
mainly consists
consists of
of the
the
shell,
shell, propulsion
propulsionsystem,
system, power
power module,
module,datadata
acquisition module,
acquisition scientific
module, sensor group,
scientific sensor
wireless Internet Internet
group, wireless of Thingsofmodule,
Things etc. The power
module, module
etc. The powerprovides
modulepowerprovidesfor functional
power for
units on Coral-AUH, while the supporting structure offers a physical
functional units on Coral-AUH, while the supporting structure offers a physical connec-connection for all
modules. The scientific sensor group is mounted in the frame and includes
tion for all modules. The scientific sensor group is mounted in the frame and includes the intelligent
pH
the (PH) electrode,
intelligent pH (PH)intelligent conductivity
electrode, intelligent(EC) electrode,(EC)
conductivity intelligent
electrode,dissolved oxygen
intelligent dis-
(DO) electrode, intelligent turbidity electrodes, self-cleaning chlorophyll
solved oxygen (DO) electrode, intelligent turbidity electrodes, self-cleaning chlorophyll digital sensors,
etc. (The
digital selection
sensors, etc.of(The
scientific sensor
selection of group wassensor
scientific determined
group according to suggestions
was determined from
according to
professional coral observation and study team in Nanhai Institute of Chinese Academy of
suggestions from professional coral observation and study team in Nanhai Institute of
Sciences.) There are two sealed cabins, which isolate inner electronic devices from water.
Chinese Academy of Sciences.) There are two sealed cabins, which isolate inner electron-
The upper sealed cabin contains a motion control module, data acquisition module and
ic devices from water. The upper sealed cabin contains a motion control module, data
wireless Internet of Things module, and the lower sealed cabin contains power supply
acquisition module and wireless Internet of Things module, and the lower sealed cabin
components such as batteries. The data acquisition control module is connected with
the scientific sensor assembly through the RS485 bus, and the data acquisition control
module can read the water quality data fed back by the sensor in real time and perform
corresponding processing by sending control signals to the water quality sensor assembly
according to the set time sequence.
 contains power supply components such as batteries. The data acquisition control mod-
ule is connected
contains with the
power supply scientific sensor
components assembly
such as through
batteries. the acquisition
The data RS485 bus, control
and themod-
data
acquisition control module can read the water quality data fed back by the sensor
ule is connected with the scientific sensor assembly through the RS485 bus, and the data in real
time and perform corresponding processing by sending control signals to the
acquisition control module can read the water quality data fed back by the sensor in real water
Sensors 2022, 22, 1770 quality
time and sensor assembly
perform according to
corresponding the set time
processing bysequence.
sending control signals to the water 3 of 15

quality sensor assembly according to the set time sequence.

Figure 2. System architecture of Coral-AUH.

Figure 2.
Figure System architecture
2. System architecture of
of Coral-AUH.
Coral-AUH.
As shown in Figure 3, the main body of Coral-AUH adopts a frame structure,
which Asisshown
As convenient
shown ininFigure
for the
Figure 3, 3,
the main
installation
the mainbody and
bodyoffixation
Coral-AUH of the
of Coral-AUH adopts
load a frame
devices,
adopts structure, which
more specifical-
a frame structure,
is convenient
ly, the is
which scientific forsensors.
convenient thefor
installation and
Theinstallation
the outer fixation
casing anofarc-shaped
andisfixation theof load
the devices,
dome
load more more
cover,
devices, specifically,
helpsthe
whichspecifical- to
scientific
improve sensors.
the The
hydrodynamic outer casing is
performance, an arc-shaped
and an dome
opening cover,
ly, the scientific sensors. The outer casing is an arc-shaped dome cover, which helps to which
channel for helps
each to improve
propeller is
the hydrodynamic
arranged on the performance,
casing to allow and
for thean opening
water to channel
flow for
through.
improve the hydrodynamic performance, and an opening channel for each propeller is each
The propeller is arranged
group con-
on thea casing
tains total to allow for the water to flow
fourthrough. Thepropellers,
propeller group contains a total
arranged on of
thesix propellers,
casing to allow including
for the water horizontal
to flow through. The which
propelleraregroup
symmetri-
con-
of six
cally propellers,
installed aroundincluding
the four
mounting horizontal
bracket propellers,
to provide which
torsion are symmetrically
force rotating installed
around the
tains a total of six propellers, including four horizontal propellers, which are symmetri-
around the
vertical mounting bracket toforce
provide torsion forcemovement.
rotating around the vertical central
cally installed around the mounting bracket to provide torsion force rotating around are
central axis and thrust for horizontal The four propellers the
axis and thrust
decoupled and force
can forindependently
horizontal movement. The forfour propellers are and
decoupled andpro-
can
vertical central axisbe and thrust force for controlled
horizontal advance,
movement. retreat
The four yaw. The
propellers are
be independently
pulsion direction controlled for advance, retreat and yaw.is theThe propulsion direction of arethe
decoupled and canofbethe remaining
independently two propellers
controlled for advance, vertical
retreatdirection.
and yaw. TheyThe pro-
remaining two installed
symmetrically propellerson is the
both vertical
sides direction.
of the They are
mounting symmetrically
bracket to provideinstalled
the on both
thrust of
pulsion direction of the remaining two propellers is the vertical direction. They are
sides
the of the mounting
vertical motion bracket
of the to provide the thrust of the vertical motion of the underwater
symmetrically installed on underwater
both sides of drone for controlling
the mounting bracket snorkeling.
to provideThis the propeller
thrust of
drone for
layout can controlling
improve snorkeling.
the space This propeller
utilization rate layout
and makecan theimprove
designedthe space
AUH utilization
structure
the vertical motion of the underwater drone for controlling snorkeling. This propeller
rate and makeand
more the designed inAUH structure more compact andstaggered
smaller inarrangement,
appearance.
layoutcompact
can improvesmaller the space appearance.
utilization rate Meanwhile,
and make through
the designed AUH structure
Meanwhile,
the flow through
channels staggered arrangement, the flow channels of each propeller do not
more compact andofsmaller
interfere with each other.
each propeller do not interfere
in appearance. Meanwhile, withthrough
each other.staggered arrangement,
the flow channels of each propeller do not interfere with each other.

Figure 3.
Figure System overview
3. System overview of
of Coral-AUH.
Coral-AUH.
Mô hình hóa động học và phân tích chuyển động
Figure 3. System overview of Coral-AUH.
2.2. Kinematic Modelling and
2.2. Kinematic Modelling and Motion
Motion Analysis
Analysis
As the
As the underwater
2.2. Kinematicunderwater
Modelling andvehicle
vehicle
Motiondesigned
designed in this
Analysisin this paper
paper is is dedicated
dedicated forfor coral
coral ecological
ecological
observation,
observation, so one key index
so one key index is to improve the athletic performance, especially its its
maneu-
As the underwater vehicleis designed
to improve in the
thisathletic
paper isperformance,
dedicated for especially ma-
coral ecological
verability, compared
neuverability, compared with the
with traditional
the torpedo-type
traditional torpedo-type autonomous
autonomous underwater
underwatervehicle.
ve-
observation, so one key index is to improve the athletic performance, especially its ma-
Considering
hicle. the requirement
Considering the for mobility,
requirement for as well as
mobility, as the
wellspecial
as the limitations,
special the layoutthe
limitations, of
neuverability,
propellers compared
was designed with the
asdesigned traditional
shown inasFigure torpedo-type
4. in
The autonomous
six propellers underwater
arepropellers
decoupledare
from ve-
each
layout
hicle. of propellers was forshown Figure 4. as
Thethe
six decou-
other, Considering
and all of themthe requirement
can independently mobility,
control theastrajectory
well of thespecial
AUH bylimitations,
changing thethe
layout of propellers
rotational speed. was designed as shown in Figure 4. The six propellers are decou-
Before establishing the kinematic model, it is necessary to define the coordinate system.
As shown in Figure 5, there are two coordinate frames: one is the ground coordinate frame
(E-X_E Y_E Z_E), also known as the static coordinate frame, which is used as the inertial
reference system, and the other is the body coordinate frame (O-xyz). In this paper, the
center of buoyancy is selected as the origin O of the coordinate frame, with the body-
referenced x-axis forward, y-axis to port (right) and z- axis down.
Sensors 2022, 22, x FOR PEER REVIEW 4 of 15

Sensors 2022, 22, 1770 pled from each other, and all of them can independently control the trajectory 4ofof the
15
AUH by changing the rotational speed.

Figure 4. Schematic diagram of the propeller layout.

Before establishing the kinematic model, it is necessary to define the coordinate sys-
tem. As shown in Figure 5, there are two coordinate frames: one is the ground coordi-
nate frame (E-X_E Y_E Z_E), also known as the static coordinate frame, which is used as
the inertial reference system, and the other is the body coordinate frame (O-xyz). In this
paper, the center of buoyancy is selected as the origin O of the coordinate frame, with
the body-referenced x-axis forward, y-axis to port (right) and z- axis down.
Figure 4.
Figure 4. Schematic
Schematic diagram
diagramof
ofthe
thepropeller
propellerlayout.
layout.

Before establishing the kinematic model, it is necessary to define the coordinate sys-
tem. As shown in Figure 5, there are two coordinate frames: one is the ground coordi-
nate frame (E-X_E Y_E Z_E), also known as the static coordinate frame, which is used as
the inertial reference system, and the other is the body coordinate frame (O-xyz). In this
paper, the center of buoyancy is selected as the origin O of the coordinate frame, with
the body-referenced x-axis forward, y-axis to port (right) and z- axis down.

Figure
Figure 5. 5. Coordinate
Coordinate frame
frame system.
system.

The
The definitionsconcerning
definitions concerningsix
sixdegrees
degreesofoffreedom
freedom (DOFs)
(DOFs) of
of Coral-AUH
Coral-AUH areare given
givenin
in the tables below, three in translations and three in rotations. The variables in theinertial
the tables below, three in translations and three in rotations. The variables in the iner-
frame
tial frameand body
and bodyframe
frameare shown
are shown inin
Tables
Tables1 and
1 and2. 2.

Table
Table 1. 1. Parameters
Parameters inin the
the inertial
inertial frame.
frame.
Parameter
Parameter Displacement
Displacement Position
Position Angle
Angle
Figure 5. Coordinate frame system.
𝑋 XE 𝜉 ξ 𝜑ϕ
𝑌Y 𝜂η 𝜃θ
The definitions
E concerning six degrees of freedom (DOFs) of Coral-AUH are given
𝑍 𝜁 𝜓
in the tables below,
ZE three in translations andζ three in rotations. The variables
ψ in the iner-
tial frame and body frame are shown in Tables 1 and 2.
Table 2. Parameters in the body frame.
Table 1. Parameters in the inertial frame.
Vector
ParameterSpeed (V) Angular Velocity (Ω)
Displacement Force (F)
PositionTorque
Angle(M)
x 𝑋 u p𝜉 X 𝜑 K
y 𝑌 v q𝜂 Y 𝜃 M
z
𝑍 w r
𝜁 Z
𝜓 N

Based on the variables defined in Tables 1 and 2, the linear velocity and angular
h . . . iT h . . . iT
velocity of Coral-AUH can be expressed as VE = ξ η ζ , VO = [u v w] T , ωE = ϕ θ ψ
 x u p X K
y v q Y M
z w r Z N

Sensors 2022, 22, 1770 Based on the variables defined in Tables 1 and 2, the linear velocity and angular 5 ofve-
15
locity of Coral-AUH can be expressed as 𝑉 = 𝜉 𝜂 𝜁 , 𝑉 = 𝑢 𝑣 𝑤 , 𝜔 =
𝜑 𝜃 𝜓 and 𝜔 = 𝑝 𝑞 𝑟 . The linear velocity and angular velocity can be converted
 . . . T
and ωO =thep coordinate
through transformation
q r . The linear velocity andmatrix [19],
angular as shown
velocity in converted
can be Equations through
(1) and the
(2).
Equation (3) is the specific expression of the speed conversion matrix 𝑅 .
coordinate transformation matrix [19], as shown in Equations (1) and (2). Equation (3) is
the specific expression of the speed conversion 𝜉 u RV
matrix
 .  𝜂 = 𝑅 v  (1)
uw
 . 𝜁
ξ
 η.  = RV  v  (1)
𝜓 ζ w
0 sec 𝜃cos 𝜑 −sec 𝜃sin 𝜑 𝑝
 .  𝜃 = 0 sin 𝜑 cos 𝜑 q (2)
𝜑 0 1 sec−tan θ cos𝜃cos
ϕ 𝜑− sec sin ϕ𝜑 r p
tanθ 𝜃sin
 
ψ
 .  
 θ = 0 sin ϕ cos ϕ  q  (2)
.
𝑅 ϕ 1 − tan θ cos ϕ tan θ sin ϕ r
 𝑐𝑜𝑠 𝜓 𝑐𝑜𝑠 𝜃 𝑐𝑜𝑠 𝜓 𝑠𝑖𝑛 𝜃 sin𝜙 − 𝑠𝑖𝑛 𝜓 𝑐𝑜𝑠 𝜙 𝑐𝑜𝑠 𝜓 𝑠𝑖𝑛 𝜃 𝑐𝑜𝑠 𝜙 + 𝑠𝑖𝑛 𝜓 sin𝜙
sin𝜙 (3)

𝑠𝑖𝑛ψ𝜓cos
= cos 𝑐𝑜𝑠θ𝜃 cos
𝑠𝑖𝑛ψ𝜓sin
𝑠𝑖𝑛θ sin φ −+sin
𝜃 sin𝜙 𝑐𝑜𝑠 𝜓 𝑐𝑜𝑠
ψ cos φ 𝜙cos𝑠𝑖𝑛 𝜓 𝑠𝑖𝑛
ψ sin 𝜃 𝑐𝑜𝑠
θ cos φ +𝜙sin− 𝑐𝑜𝑠 𝜓φ
ψ sin
RV =  sin−ψ𝑠𝑖𝑛 cos𝜃θ sin ψ sin θ sin 𝑐𝑜𝑠φ𝜃+sin𝜙cos ψ cos φ sin ψ sin θ cos 𝑐𝑜𝑠φ 𝜃−𝑐𝑜𝑠
cos𝜙ψ sin φ  (3)
− sin θ cos θ sin φ cos θ cos φ
Theoretically, six propellers can control the motion of the underwater vehicle in six
Theoretically, six propellers can control the motion of the underwater vehicle in
DOFs. However, in practical application scenarios, motion in all DOFs is not often re-
six DOFs. However, in practical application scenarios, motion in all DOFs is not often
quired. As
required. Asfor
forCoral-AUH,
Coral-AUH,the thethree
threedegrees
degreesof of freedom
freedom of of forward, steering and
forward, steering and lifting
lifting
are sufficient to meet the needs of coral observations. Based on the
are sufficient to meet the needs of coral observations. Based on the kinematic model above, kinematic model
above,
the the motion
motion analysisanalysis of Coral-AUH
of Coral-AUH is carried is out
carried out in
in these these
three three DOFs.
DOFs.
The free body diagram of the six propellers is shown in Figure 6.6.
The free body diagram of the six propellers is shown in Figure

Figure 6. Free
Figure6. Free body
body diagram.
diagram.

The
The expression
expression corresponding
corresponding to
to the
the free
free body
body diagram
diagram produced
produced by
by the
the propellers
propellers
can thus be obtained:
can thus be obtained:
X sin α1 (T1 + T2 − T3 − T4 )
   
 Y   cos α1 (T1 − T2 − T3 + T4 ) 
   
 Z   T5 + T6 
τ= 
=
   (4)
 K   L5 (− T 5 + T6 ) 

 M   0 
N L1 (−T1 + T2 − T3 + T4 )

Considering the practical application, this paper only considers three DOFs. The
force generated by the propeller can be calculated by Ti = K f Ωi 2 , where K f is the pro-
peller coefficient and Ωi is the angular velocity of the i-th propeller. Equation (4) can be
simplified as
Sensors 2022, 22, 1770 6 of 15

K f sin α1 Ω1 2 + Ω2 2 − Ω
3 2 − Ω4 2

   

X sin α1 (T1 + T2 − T3 − T4 )
τ= Z = T5 + T6 = K f Ω5 2 + Ω6 2  (5)
K f L1 Ω2 + Ω4 − Ω1 − Ω3
2 2 2 2


N L1 (−T1 + T2 − T3 + T4 )
In this paper, the Newton–Euler equation [20] is used to establish the dynamic model.
On the horizontal plane, the AUH kinetic model can be expressed as
.
Jω + ω ∗ (Jω ) + Cgyro + M f + Kζ + MC = τ (6)

where J is the total inertial of the vehicle, ω ∗ (Jω ) is the centripetal force and Cgyro is the
gyroscopic effect moment generated by the propeller due to the rotation of the propeller.
M f is water resistance or moment, which can be calculated by M f = 21 ρC f A f V 2 , where ρ
is the fluid density, C f is the water resistance coefficient, A f is the wet surface area of the
AUH and V is the flow rate. Kζ is the force and moment produced by the gravity and the
buoyancy. MC is the force and moment generated by the optical cable to the AUH, which
can be written as MC = M p + M f = 12 ρCc AC Vf 2 , where Cc is the resistance coefficient,
AC is the wet area of the cable and V f is the water velocity. τ is the force and moment
generated by the propeller.
Since the gyroscopic effect has little influence on underwater objects, it is not consid-
ered in this paper, and the resistance of the optical cable to the AUH is ignored. Combined
with the snorkeling motion in the vertical direction, the equation of motion under three
degrees of freedom can finally be simplified as
..   1 .2
mξ = K f sin α1 cos ψ Ω1 2 + Ω2 2 − Ω3 2 − Ω4 2 − ρC f H A H ξ cos ψ (7)
2
..   1 .2
mζ = −K f Ω5 2 + Ω6 2 − ρC f V AV ζ + Kζ (8)
2
..   .
JZ ψ = K f L 1 Ω 2 2 + Ω 4 2 − Ω 1 2 − Ω 3 2 − C R R ψ (9)

where C f H and C f V are the drag coefficients in the horizontal and vertical directions
obtained according to the hydrodynamic analysis, which will be introduced in detail in
Section 2.3. A H and AV respectively represent the wet area in the horizontal and vertical
directions of the AUH. CR is the general water resistance coefficient when the AUH rotates.
R is the turning radius of the designed AUH. The input control variable of the Coral-AUH
is the angular velocity of the i-th propeller. This section clarifies the dynamic law and
its quantitative expression for Coral-AUH, which prepares for the subsequent parameter
setting of the control system.
Mô phỏng CFD
2.3. CFD Simulation hiệu suất động học
In motion analysis, the water resistance is an important variable affecting the kinematic
performance of Coral-AUH, which requires accurate calculation. However, due to the
complex underwater environment, the water resistance of Coral-AUH when working
không tuyến tính và khó thu trực tiếp
underwater is not linear and can hardly be obtained directly. CFD software proves to be an
effective tool to evaluate the resistance of underwater vehicles, and the simulation results
các thông số kthuoc bên ngoài
can be used to optimize the external dimension parameters of the Coral-AUH afterward.
The diversion shell of the AUH is mainly utilized to protect the internal equipment of
the helicopter from the water flow in the marine environment and the impact damage of
various debris; meanwhile, its hydrodynamic profile helps to reduce the fluid resistance in
the operation process so as to reduce the power consumption.
In this paper, ANSYS-FLUENT is utilized as the CFD simulation tool. To ensure
calculation accuracy and save computing resources, the RNG k-ε model is used to establish
lưu vực thoát nước bên ngoài
an outer drainage basin of 4 m × 4 m × 10 m. The velocity inlet and pressure outlet
tường chống trượt
boundary conditions are established, and the rest are set as no-slip walls. Considering the
complexity of the model, unstructured grid is adopted globally. The boundary layer is
lớp ranh giới
Sensors 2022, 22, 1770 7 of 15

được tinh chỉnh gần bức tường của lưu vực bên trong và kích thước lưới tối thiểu là 2mm
refined near the wall of the inner watershed, and the minimum grid size is 2 mm. The RNG
mô hình nhiễu loạn 2 ptrinh
model is one of the widely used two-equation turbulence models; it solves two transport
equations and simulates Reynolds stress by the eddy viscosity method.
mô phỏng ứng suất R bằng pp độ nhớt xoáy
∂ρ ∂
+ (ρui ) = 0 (10)
∂t ∂xi
" !#
∂ ∂ ∂p ∂ ∂ui ∂u j 2 ∂u ∂  
− δij l −ρui0 u0j


(ρui ) + ρui u j = + µ + + (11)
∂t ∂x j ∂xi ∂x j ∂x j ∂xi 3 ∂xl ∂x j
mô hình cho động năng nhiễu loạn
The RNG k-ε model [21] is based on model transport equations for the turbulence
tốc độ tiêu tán
kinetic energy (k) and its dissipation rate (ε), which are obtained from the following
transport equations:
!
∂ ∂ ∂ ∂k
(ρk) + (ρkui ) = αk µe f f + Gk + Gb − ρε − YM + Sk (12)
∂t ∂xi ∂x j ∂x j
!
∂ ∂ ∂ ∂ε ε ε2
(ρε) + (ρεui ) = αε µe f f + C1ε ( Gk + C3ε Gb ) − C2ε ρ − Rε + Sε (13)
∂t ∂xi ∂x j ∂x j k k
sự tạo ra động năng hỗn loạn
In these equations, Gk represents the generation of turbulence kinetic energy due to
vtoc gradient trung bình
the mean velocity gradients and Gb is the generation of turbulence kinetic energy due to
sự nổi sự đóng góp của độ giãn nở dao động trong dòng chảy hỗn loạn nén vào tốc độ tiêu tán tổng thể
buoyancy. YM represents the contribution of the fluctuating dilatation in compressible
turbulence to the overall dissipation rate. C1ε , C2ε and C3ε are constants and σk and σε
số prandtl hỗn loạn
are the turbulent Prandtl numbers for k and ε, respectively. The quantities αk and αε are
hiệu quả nghịch đảo
the inverse effective Prandtl numbers for k and ε, respectively. Sk and Sε are user-defined
thuật ngữ nguồn do người dùng x/định
source terms.
The numerical solver uses the pressure-based Semi-Implicit Method for Pressure
phân tích lặp đi lặp lại
Linked Equations (SIMPLE), which iteratively decouples the calculation of velocity and
phần dư
pressure. The condition for iterative calculation to stop in each case is that the residuals of
each parameter are less than 10−3 .
Two typical translation and rotation velocities were simulated, and the resistance
obtain could be input into the motion control analysis. The resistance diagrams and the
velocity contour of the Coral-AUH are shown in Figure 7.
Comparing the resistance of the Coral-AUH in all directions under the conditions of
the incoming flow speeds of 1 and 1.5 m/s, the resistance Fx is about 25 N at 1 m/s and
dao động
about 55 N at 1.5 m/s. The resistance Fy basically fluctuates up and down around 0 N, and
the Coral-AUH receives roughly the same resistance in the x-axis and z-axis, which may
cause rising during the horizontal advance. It can be fine-tuned by the vertical propellers.
From the velocity contour, at the speeds of 1 and 1.5 m/s, the surrounding fluid flow is
similar and the streamline is good. The flow velocity of the upper fluid and the opposite
flow is relatively large and then gradually decreases to both sides of the hull, and the flow
velocity of the fluid behind the helicopter is relatively small.
hệ thống lái
In addition to translational movement, steering is also a frequent motion in underwater
moment xoắn
operations. Therefore, the torque of the Coral-AUH when it rotates at different angular
speeds perpendicular to z-axis was analyzed. The torque variations and velocity contours
tốc độ góc
when angular speed is at 90 and 180◦ /s are shown in Figure 8.
By comparison, it can be found that the torque is about 0.006 N m at 90◦ /s and 0.02 N
m at 180◦ /s. It is much smaller than the resistance encountered when moving horizontally.
It can better meet the requirements of flexible steering and maneuverability of underwater
helicopters during movement. Meanwhile, the resistance in the horizontal direction of the
underwater helicopter is small and the action is flexible, while the resistance in the vertical
direction is large, which can also ensure the stability of its navigation process.
đảm bảo sự ổn định của quá trình điều hình của nó
 Sensors
Sensors 2022,
2022, 22,22, x FOR PEER REVIEW
1770 8 of 16
8 of 15

Sensors 2022, 22, x FOR PEER REVIEW 9 of 16

Figure 7. The
Figure7. The resistance diagramand
resistance diagram andthe
thevelocity
velocitycontour.
contour.

Comparing the resistance of the Coral-AUH in all directions under the conditions of
the incoming flow speeds of 1 and 1.5 m/s, the resistance Fx is about 25 N at 1 m/s and
about 55 N at 1.5 m/s. The resistance Fy basically fluctuates up and down around 0 N,
and the Coral-AUH receives roughly the same resistance in the x-axis and z-axis, which
may cause rising during the horizontal advance. It can be fine-tuned by the vertical pro-
pellers. From the velocity contour, at the speeds of 1 and 1.5 m/s, the surrounding fluid
flow is similar and the streamline is good. The flow velocity of the upper fluid and the
opposite flow is relatively large and then gradually decreases to both sides of the hull,
and the flow velocity of the fluid behind the helicopter is relatively small.
In addition to translational movement, steering is also a frequent motion in under-
water operations. Therefore, the torque of the Coral-AUH when it rotates at different
angular speeds perpendicular to z-axis was analyzed. The torque variations and velocity
contours when angular speed is at 90 and 180°/s are shown in Figure 8.

Figure8.8.The
Figure Thetorque
torquecurves
curves and
and velocity contours.
contours.

By comparison, it can be found that the torque is about 0.006 N m at 90°/s and 0.02
N m at 180°/s. It is much smaller than the resistance encountered when moving horizon-
tally. It can better meet the requirements of flexible steering and maneuverability of un-
derwater helicopters during movement. Meanwhile, the resistance in the horizontal di-
 derwater helicopters during movement. Meanwhile, the resistance in the horizontal di-
rection of the underwater helicopter is small and the action is flexible, while the re-
sistance in the vertical direction is large, which can also ensure the stability of its naviga-
Sensors 2022, 22, 1770 9 of 15
tion process.

Hướng dẫn
3. Guidance and vàControl
phương pháp kiểm soát
Method
3. Guidance
3.1. Guidance andandActuators
Control Method
3.1. Guidance and Actuators
The positioning system of the Coral-AUH consists of an inertial measurement unit
(IMU)The andpositioning
depth sensor system of the
(Figure Coral-AUH
9). An AUH canconsists of an inertial
be equipped with anmeasurement unit
ultrashort base-
(IMU) and depth sensor (Figure 9). An AUH can be equipped with an ultrashort
line (USBL) system for underwater acoustic positioning. The USBL system is usually ex- baseline
(USBL) and
pensive system for underwater
heavy, and acoustic acoustic positioning.
positioning Thefocus
is not the USBLofsystem is usually
this paper. expensive
Relative dis-
and heavy, and acoustic positioning is not the focus of this paper. Relative
placement and angular variation data are read from IMU to locate the AUH and to veri- displacement
fyand
theangular variation
feasibility dataactuators.
of motion are read from IMU to locate the AUH and to verify the feasibility
of motion actuators.

Figure 9. Flow chart of the guidance and actuation system.

Figure 9. Flow chart of the guidance and actuation system.
The motion actuators in Coral-AUH are the horizontal and vertical propellers. The
The motion panel
motion control actuators in Coral-AUH
perceives data from aresensors
the horizontal and vertical
and outputs motion propellers.
commands The after
motion control
position panel The
calculation. perceives
inertialdata from sensors
measurement unitand outputs
(IMU) motion
measures commandsattitude
the three-axis after
position calculation.
angle and Therate
acceleration inertial
of themeasurement
Coral-AUH. The unitdepth
(IMU) measures
sensor the three-axis
measures atti-
the depth of the
tude angle and acceleration rate of the Coral-AUH. The depth sensor
AUH. Having received data from the positioning system, the motion control panel then measures the
depth of thethe
calculates AUH. Having
actual received
position data from the
of the Coral-AUH and positioning
compares system, thetarget
it with the motion con-
position
trol panel then calculates the actual position of the Coral-AUH and compares
and sends commands to the attitude adjustment system and propulsion system accordingly. it with the
target position and sends commands to the attitude adjustment system and propulsion
system accordingly.
3.2. Motion Control
The motion of the AUH requires the collaborative work of six brushless DC (BLDC)
3.2. Motion
motors. AnControl
STM32 generates PWM at a certain frequency and duty cycle, which is utilized
The motion
as drive signalsoffor
thethe
AUH requiresgroup.
propeller the collaborative
The control work of six brushless
command comes from DC (BLDC)
the data
motors. Anfrom
collected STM32 generates
sensors PWM at
and remote a certain
control frequency
signals. The IMUandmodule
duty cycle, which is of
is composed uti-
an
ITG3205
lized triaxial
as drive gyroscope,
signals an ADXL345
for the propeller triaxial
group. The accelerometer
control command and an HMX5883L
comes from thetriaxial
data
magneticfrom
collected fieldsensors
sensor.and
PIDremote
closed-loop
controlcontrol is utilized
signals. The IMU tomodule
ensure that the Coral-AUH
is composed of an
successfully
ITG3205 completes
triaxial the desired
gyroscope, motion.
an ADXL345 triaxial accelerometer and an HMX5883L triax-
In the practical
ial magnetic
Sensors 2022, 22, x FOR PEER REVIEW application
field sensor. process ofcontrol
PID closed-loop the AUH, three modes
is utilized are mainly
to ensure that theinvolved:
10Coral-
of 15
forward
AUH and backward,
successfully up and
completes the down,
desiredand steering. The operating modes of the propellers
motion.
willIn
bethe
introduced
practicalbelow (Figureprocess
application 10). of the AUH, three modes are mainly involved:
forward and backward, up and down, and steering. The operating modes of the propel-
lers will be introduced below (Figure 10).

(i) (ii) (iii)

Figure10.
Figure 10.The
Theoperating
operatingmodes
modesof
ofthe
thepropellers.
propellers. (i)
(i) forward
forward and
and backward;
backward; (ii)
(ii) up
up and
and down;
down; (iii)
(iii) left and right steering.
left and right steering.

(i) Forward and backward

When moving forward, the four horizontal propellers maintain the same rotation
speed. Propellers 1 and 4 generate a clockwise force perpendicular to the central axis,
and propellers 2 and 3 generate a counterclockwise force perpendicular to the central ax-
is. At this time, the force of the four horizontal propellers in the y-axis direction is just
Sensors 2022, 22, 1770 10 of 15

(i) Forward and backward

When moving forward, the four horizontal propellers maintain the same rotation
speed. Propellers 1 and 4 generate a clockwise force perpendicular to the central axis,
and propellers 2 and 3 generate a counterclockwise force perpendicular to the central axis.
At this time, the force of the four horizontal propellers in the y-axis direction is just offset.
The resultant force is the forward thrust. The principle of moving backward is omitted
since it is similar to that of moving forward.
(ii) Up and down
When the two vertical propellers simultaneously generate the same upward or down-
ward force according to the speed requirement, Coral-AUH moves in the vertical direction.
(iii) Left and right steering
When the horizontal propellers in relative position propel in opposite directions,
Coral-AUH can realize 360◦ turning with zero radius.

3.3. Fuzzy-PID Control Algorithm

The traditional PID controller features high reliability, simple structure and easy
parameter adjustment, but it also has some shortcomings; for example, the control accuracy
is not sufficient and the anti-interference ability is poor. The parameters in traditional PID
control are preset, and when the system is under disturbance, preset parameters cannot
be changed, so the performance of the traditional controller cannot meet certain precision.
In fuzzy algorithms, parameters can adapt to changes according to certain rules, so a
Sensors 2022, 22, x FOR PEER REVIEWfuzzy-PID control algorithm has better antidisturbance performance than a traditional 11 ofPID
15
control algorithm. The comparison of fuzzy-PID control and traditional PID control logic
diagrams is shown in Figure 11.

Figure11.
Figure Logicdiagrams
11.Logic diagramsofoftraditional
traditionalPID
PIDcontrol
controland
andfuzzy-PID
fuzzy-PIDcontrol.
control.

Thetraditional
The traditionalPID PIDcontrol
controlsystem
systemisismainly
mainlycomposed
composedofofaaPID PIDcontroller
controllerand
andaa
controlled object.
controlled object. In In this
this paper,
paper, thethecontrol
controlobject
objectisisa abrushless
brushless DC DCmotor. According
motor. Accordingto the
to
given value r ( t ) of the system and the actual output feedback c(t) of the
the given value 𝑟(𝑡) of the system and the actual output feedback c(t) of the control sys- control system, the
deviation
tem, value ofvalue
the deviation the system (t) = r (𝑒(𝑡)
of the esystem t) −=c(t𝑟(𝑡)
). − 𝑐(𝑡).

11 t 𝑑𝑒(𝑡)
 
de (t)
Z
u𝑢(𝑡)
(t) ==K𝐾p e𝑒(𝑡)(t) ++   e𝑒(𝑡)𝑑𝑡
(t)dt ++T𝑇d (14)
(14)
T𝑇i 0 d𝑑(𝑡)
(t)

p ∗∗e (𝑒(𝑡)
(t) ==K𝐾
u𝑢(𝑡) I∑
t) ++K𝐾 e(t) +
∑𝑒(𝑡) +𝐾K D [𝑒(𝑡) −𝑒(𝑡
e(t) − − 1)
e(t − 1)]k (15)
(15)
Kp K p Td
In which, K I =𝐾 =; K D; 𝐾
In which, = = . . (16)
(16)
Ti T
Fuzzy
Fuzzy control
control isis aa nonlinear
nonlinear control
control strategy
strategy based
based on
on fuzzy
fuzzy reasoning.
reasoning. Its
Its basic
basic
principle
principleisisto
tointegrate
integratehuman
humanoperation
operationexperience
experienceand
andcommon
commonsense
sensereasoning
reasoningrules
rules
through fuzzy language. When the stability of the system is destroyed, the fuzzy control
algorithm can make a judgment according to the current state to maintain stability, so
the fuzzy algorithm is integrated into the traditional PID algorithm to form a fuzzy PID
controller, which usually achieves a good control effect. As shown in Figure 11, error e
and error variation rate ec are input into the fuzzy controller, and KP’, KI’ and KD’ are
 Sensors 2022, 22, 1770 11 of 15

through fuzzy language. When the stability of the system is destroyed, the fuzzy control
algorithm can make a judgment according to the current state to maintain stability, so
the fuzzy algorithm is integrated into the traditional PID algorithm to form a fuzzy PID
controller, which usually achieves a good control effect. As shown in Figure 11, error e
and error variation rate ec are input into the fuzzy controller, and KP’, KI’ and KD’ are
obtained after calculation. These three environment-variant parameters are input into the
PID controller in real time, which facilitates the application under different circumstances.
The simulation model was established in MATLAB Simulink. In the traditional PID
control model, according to the optimization result, KP is set as 8, KI is set as 0.3 and KD
is set as 0.8. In the fuzzy PID control model, fuzzification factors Ke = 2 and Kec = 0.1;
defuzzification factors K1 = 0.8, K2 = 120 and K3 = 0.5; and initial PID parameters Kp = 0.6,
Ki = 0.8 and Ki = 0.05. The comparison of response curves in traditional PID control
and fuzzy PID control is shown in Figure 12. As shown in Figure 10, compared with the
traditional PID controller, the fuzzy PID controller has a faster response speed, shorter time
Sensors 2022, 22, x FOR PEER REVIEW
to reach the steady state and much less overshoot. When the reference input surges, 12 theof 15
response speed of fuzzy PID control is faster in the adjustment process, and there is almost
no overshoot, so the controller has a better anti-interference ability.

Figure 12.
Figure Response curves
12. Response curves of
oftraditional
traditionaland
andfuzzy
fuzzyPID
PIDcontrollers.
controllers.

4. Field Test
4. Field Test
4.1. Motion Test
4.1. Motion Test
We performed quantitative tests in a pool to measure the propulsion and attitude
We performed
adjustment capabilitiesquantitative tests in a pool
of the Coral-AUH. to measure
The average the propulsion
swimming speed in and attitude
a straight
adjustment
path was 1.54 capabilities
m/s (±0.06 ofm/s),
the Coral-AUH.
equivalent toThe average
1.5 body swimming
lengths speed
per second. Theinaverage
a straight
path
angularwasspin1.54 m/s
rate was(±0.06 m/s), (equivalent
3.14 rad/s ±0.6 rad/s)to 1.5abody
with lengthsradius.
zero turning per second. The average
The average dive
angular
speed inspin rate was 3.14
the horizontal rad/swas
attitude (±0.6 rad/s)
0.72 m/s with
(±0.07a zero
m/s),turning
equivalentradius. The
to 0.7 bodyaverage
lengthsdive
speed in the A
per second. horizontal
sample AUH attitude was 0.72
trajectory m/s (±0.07
is shown m/s),11,
in Figure equivalent to the
illustrating 0.7 body lengths
ultra-agile
swimming
per second.motion
A sample following
AUH preset commands.
trajectory is shown Thein AUH
Figure can11,change attitude,
illustrating thedirection
ultra-agile
and depth while
swimming motion exploring
following thepreset
coral reefs, with anThe
commands. average
AUHswimming
can change speed of 1.28
attitude, m/s
direction
( ± 0.26 m/s) at depths of 0 to 200 m.
and depth while exploring the coral reefs, with an average swimming speed of 1.28 m/s
(±0.26Inm/s)
order atto verify of
depths the0 feasibility
to 200 m. of the control algorithm, an experiment to test mobility
was Incarried out in a swimming
order to verify the feasibility pool.ofAthepreset route
control was designed
algorithm, and input to
an experiment into the
test mo-
control system of the AUH, and the route was quite challenging as
bility was carried out in a swimming pool. A preset route was designed and input into it had sharp turning and
attitude adjustment. In the preset task, the AUH is supposed to follow the route through
the control system of the AUH, and the route was quite challenging as it had sharp turn-
attitude adjustment and vectoring propulsion. As shown in Figure 13, the trajectory started
ing and attitude adjustment. In the preset task, the AUH is supposed to follow the route
at point O; AUH was designed to dive to Point A and then cruise to point B. After that,
through attitude◦ adjustment and vectoring propulsion. As shown in Figure 13, the tra-
AUH turned 90 clockwise and cruised to point C. At point C, AUH came up to point D,
jectory
and then started at point
it cruised O; AUH
to point E. At was
pointdesigned to dive90
E, AUH turned to◦ Point A and
clockwise andthen cruise
finally to point
returned
B. After that, AUH turned 90° clockwise and cruised to point
to point O. The whole route mainly involves variations in yaw. A comparison was made C. At point C, AUH came
up to point
between theD, and then
preset routeitand
cruised to point
the actual E. At
route, andpoint E, AUH
the result turned
is shown in90° clockwise
Figure 13. Yawand
finally
angle wasreturned to point in
also recorded O.real
Thetime
wholeandroute mainlywith
compared involves variations
the preset value,inasyaw.
shown A com-
in
parison was made between the preset route and the actual route, and the result is shown
in Figure 13. Yaw angle was also recorded in real time and compared with the preset
value, as shown in Figure 14. The actual route and heading angle agree well with preset
values, validating the fuzzy control algorithm.
 through attitude adjustment and vectoring propulsion. As shown in Figure 13, the tra-
jectory started at point O; AUH was designed to dive to Point A and then cruise to point
B. After that, AUH turned 90° clockwise and cruised to point C. At point C, AUH came
up to point D, and then it cruised to point E. At point E, AUH turned 90° clockwise and
Sensors 2022, 22, 1770 finally returned to point O. The whole route mainly involves variations in yaw. 12 of 15A com-
parison was made between the preset route and the actual route, and the result is shown
in Figure 13. Yaw angle was also recorded in real time and compared with the preset
value,
Figureas
14.shown in Figure
The actual 14. heading
route and The actual route
angle agreeand
wellheading angle
with preset agree
values, well with
validating the preset
values, validating
fuzzy control the fuzzy control algorithm.
algorithm.

Sensors 2022, 22, x FOR PEER REVIEW 13

Sensors 2022, 22, x FOR PEER REVIEW 13 of 15

Figure13.
Figure The 3D
13. The 3D trajectory
trajectoryofofCoral-AUH.
Coral-AUH.

Figure 14. Heading angle of Coral-AUH (yaw).

Figure14.
Figure 14.Heading
Heading angle
angle of Coral-AUH
of Coral-AUH (yaw).(yaw).

4.2. Field Test

4.2.
4.2. Field
FieldTest
Test
Thefunction
The functionandandreliability
reliabilityof of Coral-AUH
Coral-AUH were
were tested
tested in ain a field
field test
test in inwaters
the the waters
of of
The function
Luhuitou and andBay
Yazhou reliability
in Sanya ofCity.
Coral-AUH
Coral were
reef testedrestoration
ecological in a field test inlocated
areas the wate
Luhuitou and Yazhou Bay in Sanya City. Coral reef ecological restoration areas located in
Luhuitou
in the
the andNational
Nanhaisuo
Nanhaisuo Yazhou
National Bay
Coral in Sanya
Coral City.
AreaCoral
Conservation
Conservation Area
and reef
theand ecological
Islandrestoration
the Dongluo
Dongluo Island
Coral Coralareas
Restora- Res-loc
in
tionthe Nanhaisuo
toration
Area Area wereNational
were selected selected Coral
as test as test
sites Conservation
in FigureinArea
sites (shown
(shown Inand
Figure
15). the
15).
the InDongluo
field the field
test, Island
test,
Coral-AUH Coral
Coral-
AUH successfully
toration Area
successfully were
cruised cruised
to to the
selected
the desired as desired
site test site(shown
with sites
a presetwith
route.a The
preset route.
in Figure
photos The
15).
taken Inphotos
by takentest,
the field
Coral-AUH by Co
Coral-AUH
are shown in are shown
Figures 16 in
and Figures
17, and 16
the and 17,
water and
qualitythe
datawater quality
collected
AUH successfully cruised to the desired site with a preset route. The photos taken by data
the collected
sensor groupby the
are shown
sensor
Coral-AUH in Tables
group are 3 and 4.
are shown
shown in
inTables
Figures 3 and
16 4.
and 17, and the water quality data collected by
sensor group are shown in Tables 3 and 4.

Figure15.
Figure 15.Coral-AUH
Coral-AUHinin field
field test.
test.

Figure 15. Coral-AUH in field test.

Sensors 2022, 22, 1770 13 of 15

Figure 15. Coral-AUH in field test.

Figure 15. Coral-AUH in field test.

Figure 16.National
National CoralConservation
Conservation Areain
in Luhuitou.
Figure 16. National Coral
Figure 16. Coral Conservation Area
Area in Luhuitou.
Luhuitou.

Figure 17. Picture taken under the sea of Dongluo Island.

Figure 17.
Figure 17. Picture
Picture taken
taken under
under the
the sea
sea of
of Dongluo
Dongluo Island.
Island.

Table 3. Water quality data of National Coral Conservation Area.

Position
Electric Conductivity Dissolved Oxygen Turbidity Chlorophyll
Eastern Northern pH
(uS/cm) Concentration (mg/L) (FTU) (ug/L)
Longitude Latitude
18◦ 120 4200 109◦ 280 2100 8.08 47.8 6.32 21.05 3.02
18◦ 120 4500 109◦ 280 2500 8.13 46.35 6.17 19.83 3
18◦ 120 4600 109◦ 280 3400 8.11 47.48 5.93 23.42 2.72
18◦ 120 4700 109◦ 280 3800 8.13 52.64 5.89 16.8 3.39

Table 4. Water quality data of Dongluo Island.

Position
Electric Conductivity Dissolved Oxygen Turbidity Chlorophyll
Eastern Northern pH
(uS/cm) Concentration (mg/L) (FTU) (ug/L)
Longitude Latitude
18◦ 190 3200 108◦ 590 3000 8.06 47.8 5.99 20.21 3.2
18◦ 190 2800 108◦ 590 2000 8.15 46.35 6.03 35.02 4.81
18◦ 190 4200 108◦ 590 1800 8.2 47.48 6.01 15.98 2.91
18◦ 190 4400 108◦ 590 1900 8.25 52.64 5.91 16.13 3.23

The field test lasted 2 h each time. The maximum depth reached was 30 m. It can be
seen from the test results that the Coral-AUH successfully cruised to the coral reef site and
hovered there to take photographs; meanwhile, it could detect environmental data such as
pH, conductivity, dissolved oxygen, turbidity and chlorophyll in real time. The test results
show that the system can improve the accuracy and timeliness of coral monitoring and
can meet the long-term stereoscopic observation requirements for coral reefs in a complex
seabed ecological environment.

5. Conclusions
This paper contributes to the field of undersea robots that can serve as an observatory
platform for coral reefs. First, we present the design considerations of a stingray-shaped
underwater vehicle that can complete scientific surveys continuously within certain sea
areas. Motion analysis and CFD simulation show Coral-AUH has low rotation resistance
and superior maneuverability. A fuzzy algorithm is integrated into the traditional PID
Sensors 2022, 22, 1770 14 of 15

algorithm to form a fuzzy PID controller, which usually achieves a good antidisturbance
effect. Simulation and experimental results show that AUH follows the preset route well
and demonstrates excellent maneuverability. Finally, the function and reliability of Coral-
AUH were tested in the waters of Luhuitou and Yazhou Bay at the depths of 10 and 30 m,
respectively, and the duration of each test lasted for around 2 h. The results show that
Coral-AUH can quickly, accurately and efficiently conduct coral observation tasks.

Author Contributions: Conceptualization, H.H. and Y.C. (Ying Chen); motion analysis, N.Z.; CFD
analysis, L.Z.; hardware, Y.C. (Yuchao Che) and J.G.; field test, J.Z., Y.C. (Yuchao Che) and L.Z.;
writing—original draft preparation, N.Z.; writing—review and editing, J.Z.; supervision, J.Z. All
authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by Hainan Key R&D Project (ZDYF2020197), Fundamental
Research Funds for the Central Universities (2021XZZX014).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: All the data supporting the reported results has been included in
this paper.
Acknowledgments: We wish to thank the timely help given by Jin Guo in CFD simulation analysis.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Burke, L.; Reytar, K.; Spalding, M.D. Reefs at risk revisited in the coral triangle. Ethics Med. 2011, 22, 2008–2010.
2. Xing, X.; Wang, G.; Li, M.; Chen, Y.; Tian, T. Development of a comprehensive monitoring system on environmental information
in sea ranching. J. Dalian Ocean. Univ. 2017, 32, 105–110.
3. Xavier, H.; Madhumi, M.; Abhijit, N. Using remote sensing and remotely operated vehicles to improve ground-truthing efforts
for underwater resources in a developing island nation. J. Appl. Remote Sens. 2013, 1, 1–8.
4. Kilgour, M.J.; Auster, P.J.; Packer, D.; Purcell, M.; Packard, G.; Dessner, M.; Rissolo, D. Use of AUVs to inform management of
deep-sea corals. Mar. Technol. Soc. J. 2014, 48, 21–27. [CrossRef]
5. Morris, K.J.; Bett, B.J.; Durden, J.M.; Huvenne, V.A.; Milligan, R.; Jones, D.O.; Ruhl, H.A. A new method for ecological surveying
of the abyss using autonomous underwater vehicle photography. Limnol. Oceanogr. Methods 2014, 12, 795–809. [CrossRef]
6. Peng, W. The Research of Underwater Vehicle Used on Marine Ranching; Shanghai Ocean University: Shanghai, China, 2014.
7. Berget, G.E.; Fossum, T.O.; Johansen, T.A.; Eidsvik, J.; Rajan, K. Adaptive Sampling of Ocean Processes Using an AUV with a
Gaussian Proxy Model. IFAC-PapersOnLine 2018, 51, 238–243. [CrossRef]
8. Wynn, R.B.; Huvenne, V.A.I.; Le Bas, T.P.; Murton, B.J.; Connelly, D.P.; Bett, B.J.; Ruhl, H.A.; Morris, K.J.; Peakall, J.;
Parsons, D.R.; et al. Autonomous Underwater Vehicles (AUVs): Their past, present and future contributions to the advancement
of marine geoscience. J. Mar. Geol. 2014, 352, 451–468. [CrossRef]
9. Doya, C.; Chatzievangelou, D.; Bahamon, N.; Purser, A.; de Leo, F.C.; Juniper, S.K.; Thomsen, L.; Aguzzi, J. Seasonal monitoring of
deep-sea megabenthos in Barkley Canyon cold seep by internet operated vehicle (IOV). PLoS ONE 2017, 12, e0176917. [CrossRef]
[PubMed]
10. Laschi, C.; Mazzolai, B.; Cianchetti, M. Soft robotics: Technologies and systems pushing the boundaries of robot abilities. Sci. Robot.
2016, 1, eaah3690. [CrossRef] [PubMed]
11. Yoshida, H.; Aoki, T.; Osawa, H.; Ishibashi, S.; Watanabe, Y.; Tahara, J.; Miyazaki, T.; Itoh, K. A deepest depth ROV for sediment
sampling and its sea trial result. In 2007 Symposium on Underwater Technology and Workshop on Scientific Use of Submaine Cables and
Related Technologies; IEEE: Piscataway, NJ, USA, 2007; pp. 28–33.
12. Chen, C.-W.; Jiang, Y.; Huang, H.-C.; Ji, D.-X.; Sun, G.-Q.; Yu, Z.; Chen, Y. Computational fluid dynamics study of the motion
stability of an autonomous underwater helicopter. Ocean Eng. 2017, 143, 227–239. [CrossRef]
13. Feng, Z.; Chen, C.-W.; Yu, Z.; Huang, H.-C.; Su, H.; Leng, J.-X.; Chen, Y. Hydrodynamic Performance Analysis of Underwater
Helicopter Based on STAR-CCM+. In OCEANS-MTS; IEEE: Piscataway, NJ, USA, 2018.
14. Cai, Q.-W.; Chen, C.-W.; Yu, Z.; Huang, H.-C.; Yan, N.-M.; Su, H.; Leng, J.-X.; Chen, Y. Numerical Analysis of Hydrodynamic
Interaction for AUH and Mother Ship in Different Regular Wave Directions. In OCEANS-MTS; IEEE: Piscataway, NJ, USA, 2018.
15. Chen, C.-W.; Yan, N.-M. Prediction of added mass for an autonomous underwater vehicle moving near sea bottom using panel
method. In Proceedings of the 4th International Conference on Information Science and Control Engineering (ICISCE 2017),
Changsha, China, 21–23 July 2017; pp. 1094–1098.
16. Chen, C.; Yan, N.; Leng, J.; Chen, Y. Numerical analysis of second-order wave forces acting on an autonomous underwater
helicopter using panel method. In OCEANS-MTS; IEEE: Piscataway, NJ, USA, 2017.
Sensors 2022, 22, 1770 15 of 15

17. Wang, Z.; Liu, X.; Huang, H.; Chen, Y. Development of an underwater autonomous helicopter with high maneuverability.
Appl. Sci. 2019, 9, 4072. [CrossRef]
18. Ji, D.; Chen, C.; Chen, Y. Autonomous Underwater Helicopters, AUV with Disc-shaped Design for Deepwater Agility. Sea Technol.
2018, 59, 25–27.
19. Jiang, X. Unmanned Underwater Vehicles; Liaoning Science and Techinology Publishing House: Shenyang, China, 2000; pp. 244–248.
20. Titterton, D.; Weston, J.L.; Weston, J. Strapdown Inertial Navigation Technology, 2nd ed.; The Institution of Engineering and
Technology: London, UK, 2004; p. 574.
21. Yakhot, V.; Orszag, S.A. Renormalization-Group Analysis of Turbulence. Phys. Rev. Lett. 1986, 57, 1722–1724. [CrossRef]
[PubMed]