Design and Locomotion Control of a Myliobatid-inspired Robot

Actuated by Passively-flexing Pectoral Fins

Songzi Guo

, Zhiyin Li

and Jinhua Zhang

China Ship Design and Development Center, Wu’han 430034, China

Key Laboratory of Education, Ministry for Modern Design and Rotor-Bearing System, Xi’an Jiaotong University,

Xi’an 710049, China

Keywords: Bionic Design, Robot Fish, Flexible Pectoral Fin, Bio-Inspire.

Abstract: This article proposes the mechanical design of a myliobatid-inspired robot (XJRoman-Ⅰ) based on oscillatory

swimming mechanism for both stability and agile manoeuvrability. Inspired by anatomical studies, a pair of

passively bending pectoral fins are developed to generate propulsive force for the prototype. An elevator is

adopted to adjust its pitch attitude. Primary experimental research on the effect of fin’s spanwise stiffness on

swimming performance is performed to improve its swimming performance. By embedding a stiff rod into

the fin’s leading edge, the thrust and lateral force generated by the fins are improved significantly. Finally, a

CPG-based control method is introduced to make the prototype achieve different locomotion patterns

including cruising by flapping pectoral fins and turning by modulating phase relation of pectoral fins. This

paper mainly focuses on propulsive capability of stability and agility for the prototype, and expects to propose

an excellent underwater vehicle covering wide range of applications.

1 INTRODUCTION

Traditional underwater vehicles play an important

role in underwater explorations. But with the growing

demands for pelagic and deep-sea missions, they

couldn’t well adapt to these complex and

challengeable operating scenarios because of

limitation of their propulsive mechanisms. Hence, in

order to solve this conundrum, biologists and

engineers attempt to mimic propulsive mechanisms

of marine animals to enhance the performance of

underwater vehicles. Admittedly, marine animals are

quite different from underwater vehicles in many

aspects. Yet even so, by understanding how aquatic

animals achieve these inborn advantages, engineers

can be inspired to develop a variety of novel undersea

robots like a robotic dolphin (Yu, 2017), knife-fish

(Liu, 2017) and trout (Takemura, 2011) giving man’s

horizon further into the ocean than ever before.

Among the numerous aquatic animals, myliobatid,

renowned as a large open-water swimmer, has

streamlined body shape and peculiar propulsive

mechanism (Rosenberger, 2001). While the majority

of fishes use their caudal fin to generate propulsion

and pectoral fins to keep balance, rays synchronically

or asynchronously flap their flexible pectoral fins on

each side of the body to achieve agile manoeuvre and

stable locomotion, which take great advantages over

other caudal fin-based fishes in long-distance

swimming (Breder,1926) and agile motion

(Fish,2018).

Researchers intensively investigated kinematics

characteristics of myliobatid fishes, attempting to

reveal the mechanism behind such great

comprehensive abilities of high efficiency and

stability during cruise. By analysing the typical

frames of a cruising manta ray, Liu et al (Liu, 2015)

established a kinematics model to delineate

deformation of pectoral fins in both span and chord.

The model indicated that the deformation in the span

becomes significantly at distal part of pectoral fins.

The flexible deformation in span was proved to allow

mantas to perform small radius turns with its ratio as

high as 67 deg/s (Fish et al., 2018), despite their rigid

body (Parson, 2011). Based on the kinematics

research on mantas, various computational

hydrodynamic simulations (Fish et al., 2016) were

carried out and indicated that the flexible distal part

of pectoral fins plays an essential role in mantas’

swimming and generates the majority of thrust. On

top of this, comprehensive experimental studies were

carried out by Clark and Smits (Clark et al, 2006).

Guo, S., Li, Z. and Zhang, J.

Design and Locomotion Control of a Myliobatid-inspired Robot Actuated by Passively-ﬂexing Pectoral Fins.

DOI: 10.5220/0010857000003271

In Proceedings of the 19th International Conference on Informatics in Control, Automation and Robotics (ICINCO 2022), pages 353-360

ISBN: 978-989-758-585-2; ISSN: 2184-2809

353

They found the efficiency of a flapping foil is

promoted by flexibility distributed on the foil. With

the help of DPIV technology, Dewey et al (Dewey et

al., 2012; 2013) observed wake topology formed by a

bionic pectoral fin and proposed a resonance theory,

which indicated a maximal efficiency is always

achieved while flapping pectoral fins is actuated

under its resonance frequencies.

By mimicking the kinematics features of

oscillatory batoid fishes, many newly-designed

underwater robots have been developed. In 2011, a

ray-like robot fish was initially design and fabricated

by Low et al (Low, 2012), whose pectoral fin in one

side was actuated by three servos in parallel.

Similarly, this parallel connected actuating system

was also utilized by an improved ray-like robot (Cao,

2019). A cpg-fuzzy-based control method was

introduced to better mimic the swimming gaits of

myliobatid. However, the robot fish can actively

control multiple degrees of freedom along the chord

by means of groups of servos, redundant actuators

lead to extra loss of mechanical energy and limit

swimming performance of robots. Therefore, Chew et

al (Chew, 2016) have developed a manta-like robot

merely actuated by two servos. Because of the

lightweight and compact design, maximal swimming

speed of this robot fish can reach to 1.7 body length

per second.

In this paper, the design of a myliobatid-inspired

robot equipped with a pair of passively bending

pectoral fins based on oscillatory swimming

mechanism is proposed. The fin’s design is based on

the anatomical structure of an eagle-ray. A traveling

wave alone the chord, coupling with a vertical wave

alone the wingspan is presented on the flexible fin

during flapping in the water. In addition, an

experiment on the effect of span-wise stiffness is

carried out to improve the swimming performance of

the prototype. Finally, the prototype equipped with

the optimized pectoral fins is tested in a pool to verify

its capability of locomotion.

2 FROM BIOMIMETIC SUBJECT

TO BIOINSPIRED ROBOT

2.1 Overview of the Prototype

To mimic stable and flexible swimming mode of

myliobatid, a bionic oscillatory robot equipped with a

pair of flexible pectoral fins is constructed, as shown

in Figure1(a). The robot prototype has a flat and

streamlined body shape suitable to achieve both fast

and stable motions in the water and equip with a pair

of flexible pectoral fins, and can be divided into four

parts: (1) a head cabinet fixing two of actuating servos

(Omg, 40kgf.cm) providing torque force for the

bionic fins; (2) a central waterproof shell containing

electric components; (3) an elevator to adjust pitch

attitude; (4) a pair of flexible pectoral fins providing

propulsive force for the prototype. The head cabinet

as well as the central shell are made of the lightweight

material (nylon-12), whose density is close to water.

The weight of the robot is 680g, and it can provide an

extra load capacity of 200g. A central controller

(RoboMaster, DJI) powered by FreeRTOS system is

adopted to control locomotion of the prototype.

Considering that the effect of attenuation for signals

will be intensive with the increasing frequency in the

water, a radio module with lower transmission

frequency is adopted as communication unit.

Additionally, the central waterproof shell

accommodates sensors for underwater environment

perception. An inertial measurement unit (IMU)

embedded in the controller is arranged in parallel with

the robot principle axis to sense the three-dimensional

attitude for the prototype with a sampling rate of 50

Hz. A developed pressure sensor MS5837 is mounted

at the end of the shell to gauge hydraulic pressure

corresponding to the depth where the robot stays. It

has a 0.2mbar resolution in a scale range of 0-30bar.

The elevator at the stern of the prototype driven by a

servo performs vertical rotation to provide pitch

torque for the prototype.

Figure 1: (a) Isometric view of the robotic prototype. (b)

Components of the robot prototype.

ICINCO 2022 - 19th International Conference on Informatics in Control, Automation and Robotics

354

Table 1: Main Technical Specifications of the Robotic

Prototype.

Items Characteristics

Dimension

385mm

×560mm×50mm

Total mass 680g

Control unit STM32F407IGH

Sensors IMU(BMI088), Pressure sensor

Drive units

Servo

×3(waterproof)

Communication unit AS32170-170MHz

Power supply 11.1V(2200mAh-Li)

2.2 Design and Fabrication of Bionic

Pectoral Fins

The design of the bionic pectoral fin is based on the

anatomical details of a myliobatid swimmer, eagle-

ray. It is a typical swimmer using oscillatory model

embodying an excellent balance between

manoeuvrability and efficiency. For this reason, their

anatomical characteristics would inspire the design of

next-generation underwater vehicles. To uncover the

outstanding swimming performance of eagle-rays, we

conduced anatomical experiments on its pectoral fins

to discover the smallest detail in musculoskeletal

structure, which is shown in Figure2.

Figure 2: (a) The fin’s skeletal structure of an eagle-ray.

(b) The cross-section of the dissected pectoral fin.

The pectoral fin of the eagle-ray is approximately

200 mm in chord and 220 mm in span, where span is

measured from the fin tip to the root. The fin’s aspect-

ratio is close to 2.2 that is the ratio of the span-wise

length to the fin’s surficial area.

After removing tissues from the specimen, skeletal

structure can be clearly observed in Figure2(a). The

pectoral fin consists of an array of 175 fin-rays. Each

of the fin-ray is composed by several segments

connected by cartilaginous joints. These segments

extend out from the root to the edge of the pectoral fin

and arrange in radial formation. In Figure2(b), the

thickness of the pectoral fin decreases from the basal

part to the distal part. This kind of skeletal and

muscular structure is considered to be favourable for

the range of motion for fins tip, by which agility and

efficiency of eagle-rays can be enhanced.

Additionally, skeletal connections also exist between

the adjacent radial segments. This connective tissue

is termed cross-bracing, which can limit the

oscillatory amplitude of adjacent fin-rays and

increase the chord-wise stiffness in the pectoral fin

necessary for the transmission of traveling wave

along the chord-wise direction.

Based on the morphologic characteristics and

anatomical details of the specimen, the design of a

bionic pectoral fin to mimic the oscillatory propulsion

mechanism is proposed. The chord-length and span-

length of the bionic fin are 220mm and 200mm,

respectively. The AR of the bionic fin is about 2.2.

The leading edge that has the longest length

among fin-rays connects with a servo (Omg,

40kgf.cm), providing oscillatory force for the

pectoral fin. The rest of the fin-rays at the middle and

the end of the bionic pectoral fin respectively can

passively control the wavelength alone the chord

presenting on the pectoral fin. To make the bionic

pectoral fin achieve passive deformation alone the

wingspan and allow the flexible membranes to

maintain it shape while flapping in the water, all of

the fin-rays are made of nylon with high-ductility.

The thickness of the fin-rays decreases from the basal

part to the distal part. Therefore, the tip of the pectoral

fin can achieve more significant deformation than

proximal part. Two of the flexible membranes are

made of silicon rubber and attached to the fin-rays by

adhesive (E41, Wacker). Since M4601, a kind of two-

component silicone rubber, has appropriate tensile

strength (6.5N/mm

) and lower hardness (28A)

compared with other silicon rubbers, it is adopted to

fabricate flexible membranes for the bionic pectoral

fins. Firstly, the unmixed M4601-A and M4601-B are

poured into a container by weight of 1:9. After

degassing in a vacuum chamber, the mixed material

is poured into resin moulds where the demould

releaser has been sprayed. And then, after curing at

room temperature for 12 hours, the vulcanized

specimen is stripped from the moulds. The cured fin-

like membranes have 2mm in thickness. Finally,

Design and Locomotion Control of a Myliobatid-inspired Robot Actuated by Passively-ﬂexing Pectoral Fins

355

when all of the compliant membranes are fabricated.

They are adhered to the compliant fin-rays. The

fabricated bionic pectoral fin is shown in Figure3.

Figure 3: The fabricated bionic pectoral fin, noted that in

this paper the bionic fins are dyed into different colours by

fluorescent agent to make them more distinguishable.

2.3 Experiment on Bionic Pectoral Fins

In order to improve propulsive performance for the

robot, experimental research was carried out to

investigate the effect of stiffness distribution along

the leading edge on thrust and lateral force generated

by the pectoral fins. On top of that, the position of the

passive bending along the leading edge is investigated,

since it determines the proportion of the distal part on

the bionic fin, which has been proved to be closely

relative to thrust generation in Fish’s work (Fish,

2016).

In the experiment, we studied five bionic fin’s

configurations whose leading edge is different in the

proportion of high stiffness part marked with a red

bracket in Figure4(a). An aluminium rod is embedded

into the rear side of fin’s leading edge to form discrete

region with high stiffness along the span-wise

direction. The fabricated bionic fins are demonstrated

in Figure4(b) where the proportion of the high

stiffness part decreases from 1 ( the pink pectoral fin)

to 0 (dull-red pectoral fin). In the fins’ naming

scheme, ‘HS-*’ refers to the proportion of the leading

edge with high stiffness part.S

Figure 4: (a) The leading edges embedded with different

length of aluminium rods. (b) The five fabricated bionic fins

with different scale of high stiffness region.

The experimental apparatus is shown in Figure5.

A servo clamped by a carriage was used to actuate the

bionic fins. The flapping apparatus is mounted on the

interface of a 6-axis force torque sensor (ATI-Mini40,

0~20N) and submerged in a water tank. Since the

passive bending appearing on the leading edge is

critical to the thrust generation, a high-speed camera

(Phantom, v1612) was placed in the front of the water

tank to capture the frames to investigate the passive

bending appearing on the leading edge during

flapping.

Figure 5: Annotated diagram of the experimental apparatus.

Myliobatid prefers to change the flapping

frequency to meet different motion requirements

(Fish, 2016). Therefore, the five bionic fins were

actuated under different stroke frequencies from 0.3

to 1.0 Hz with a constant flapping amplitude of 80°.

To make the experimental data acquired in our work

reliable, the collecting duration of each set of data

were lasting for 20 flapping circles. The thrust and

lateral force throughout ten stable flapping circles

were selected to obtain the averaged thrust and lateral

force.

Cycle-averaged thrust and lateral force generated

by the bionic fins are shown in Figure6(a) and

Figure6(b), respectively. Compared with the bionic

fin’s configuration HS-0 (Dull-red) without stiffener

in the leading edge, the other fin’s configurations

have overall increases in averaged thrust and lateral

force. Additionally, except for the bionic fin termed

HS-1 (Pink), the averaged thrust increases with the

increasing stoke frequency before a turning point,

where the averaged thrust reaches its maximum and

then starts to decrease. On the other hand, the

averaged lateral force demonstrates incremental trend

with the increasing stroke frequency. The fin’s

configuration HS-1 embedded with a stiff rod as long

as its leading edge can achieve the highest cycle-

averaged thrust among the fin’s design, which is

1.32N at stroke frequency of 1.0 Hz. However, its

generated lateral force can’t be ignored. Because a

pair of bionic pectoral fins mounted on the robot

ICINCO 2022 - 19th International Conference on Informatics in Control, Automation and Robotics

356

prototype couldn’t be perfectly symmetrical in size,

the discrepancy in lateral force make the prototype

difficult to keep its course. If the fundamental need

for the robot prototype is to perform cursing

locomotion that focuses on the swimming speed and

stability, the ideal pectoral fins are expected to

provide a significant thrust but relatively unobvious

magnitude of lateral force to maintain the swimming

stability. Based on the experimental result, the HS-0.5

(blue) is chosen as an ideal actuator for the robot

prototype, as it can achieve great averaged forces

even closer to that of the HS-1 at stroke frequencies

from 0.3 Hz to 0.7Hz while the averaged lateral forces

are relatively lower than other pectoral fins.

Figure 6: (a) The cycle-averaged thrust and (b) the lateral

force at different stroke frequencies from 0.3 Hz to 1.0 Hz.

The hollow symbol demonstrates the stroke frequency

climbs to the turning point, where the averaged thrust

begins to drop down for the next stroke frequency.

While a bionic fin is flapping, the position of a

passively-bending point along the fin’s leading edge

marked with a couple of red arrows in Figure7 can

serve as an indicator for the magnitude of

hydrodynamic load on the bionic fin, which is relative

to the force generation of the bionic fin. Therefore,

the image data acquired from the high-speed camera

were analysed to obtain the position of the passively-

bending points.

Figure 7: The snapshots of the different designs of bionic

fins while actuating at stroke frequency of 1 Hz. The

passively-bending points are mark with red arrows.

Figure 8: The proximal stiff part occurring on the bionic fin.

The hollow symbol demonstrates the stroke frequency

where the variation tendency in the chart is changed.

The proportion of the proximal stiff part occurring

on the bionic fin is shown in Figure8. For the fin’s

configuration of HS-1.0 (Pink), the passive bending

didn’t appear on its leading edge, as the leading edge

of the fin is rigid enough to resist the reactive force.

In contrast, the proximal stiff part of the other bionic

fins decreases with the stroke frequency increasing

because of the increasing hydrodynamic load on the

fin. However, as the stroke frequency further

increases to a restraint frequency, the proportion of

the proximal part begins to keep a constant value. It

is because the embedded aluminium rod commences

to restrain the passively-bending point from shifting

towards proximal part further,

It should be noted that compared with Figure6(a)

and Figure8, the bionic fins reach their maximum in

averaged thrust as the proportion of the proximal part

just drop down to its bottom. For the next stroke

frequency, the generated thrust begins to decrease.

We deduce that the distal part on the bionic fin ploys

an important role in thrust generation. While the

stroke frequency increases from 0.3 Hz to the trans-

Design and Locomotion Control of a Myliobatid-inspired Robot Actuated by Passively-ﬂexing Pectoral Fins

357

frequency, the increment in thrust takes advantage

over the increment in drag force. Hence, the gross

magnitude of the thrust shows a growing trend. On

the contrary, while stroke frequency increases beyond

the trans-frequency, the increment in drag force takes

advantage over the increment in thrust, since the scale

that can generate thrust is restrained.

3 LOCOMOTION CONTROL

BASED ON A CPG-BASED

CONTROL METHOD

Owing to central pattern generators (CPGs)

modulating the rhythmic movements, animals can

perform various locomotion patterns with excellent

stability. They can be also employed to control

coordination locomotion of robots with multi-degree

of freedom especially in performing swimming

patterns of robotic fish and other types of robots

(Zhou and Low, 2012). Many sorts of the CPG

models have been employed to control locomotion for

bio-inspired robots, such as Hopf model (Zhou and

Low, 2010a), Matsuoka’s model (Matsuoka, 1985)

and Ijspeert’s model (Ijspeert et al., 2007). In this

paper, we adopt a simplified linear CPG model

proposed by (Wang et al., 2017) to generate rhythmic

signals for three servomotors corresponding to the

fins and the tail. The CPG controller shown in Figure

9 is composed of three coupled oscillators and it is

implemented as follows:

(-)

2(--)

cos( )

iiii

i i ij j i ij

iii i

aAa

bBb

xf xx

ba x

μϕ

∈















(1)

Figure 9: The CPG network of XJRoman-Ⅰ.

The parameters a

, b

and x

are state variables of the

CPG model and represent the amplitude, offset and

phase of the rhythmic signal produced by the ith

oscillator. A

, B

and f

are control parameters to

modulate the desired flapping amplitude, offset and

frequency. φ

is a parameter for the desired phase

relation between the ith oscillator and jth oscillator.

is a coefficient denoting the coupling strength

between the ith oscillator and jth oscillator. η

and β

are converging coefficient affecting the convergence

speed of the CPG network. The output θ

is the input

angle of the servomotor. The subscripts i=1, 2, 3

correspond to the left pectoral fin, right pectoral fin

and the tail of the prototype, respectively

4 SWIMMING TEST FOR THE

PROTOTYPE

In order to acquire cruise speed of the prototype, we

tested the prototype equipped with the blue pectoral

fins (HS-0.5) in a swimming pool under stroke

frequencies from 0.5 Hz to 1.2 Hz with a fixed

flapping amplitude (80°). In cruise pattern (Figure10),

CPG1 and CPG2 modulate synchronous rhythm

signals to actuate pectoral fins to keep the robot

swimming along the straight course. While actuated

at the a stroke frequency of 0.8 Hz, the prototype can

reach its maximal cruising speed at 1.8 BL/s (body

length per second).

Figure 10: Output signals of the CPGs while the prototype

performing cruise locomotion.

Table 2: Parameter Values in CPGs (Cruise Pattern).

[1,1,1] 0 [60,60,k] [0,0,0] [1,1,1] [1,1,1] 1

In turning pattern, The modulated signals are

recorded in Figure11, where the CPG1 and CPG2

modulate asynchronous rhythm signals to actuate the

pectoral fins asynchronously to alter swimming

course of the robot gradually. Based on the snap shots

of the turning pattern, the CPG-based controller can

make the prototype alter swimming gaits stably.

ICINCO 2022 - 19th International Conference on Informatics in Control, Automation and Robotics

358

Figure 11: Output signals of the CPGs while the prototype

performs turning gait.

Table 3: Parameter Values in CPGs (TurningPattern).

[1,1.2,1]

[60,60,k] [0,0,0] [1,1,1] [1,1,1]

5 POTENTIAL ONGOING WORK

In further work, we are going to investigate

propulsive efficiency of the robot prototype, since it

is a critical parameter to evaluate the swimming

performance of the prototype, especially in fulfilling

long-distance mission.

On the other hand, since the flapping bionic fin

can also provide lift force for the robot, it can be used

to achieve gliding locomotion for the prototype to

save energy during swimming. In our further work,

we also plan to investigate the gliding locomotion for

the prototype to make it fulfil a long-distance travel.

6 CONCLUSION

In this study, a prototype of oscillatory robot based on

the combination of biological swimming mechanisms

and morphological features was proposed to make the

prototype achieve both fast and stable locomotion

through a pair of bionic pectoral fins. Moreover,

primary research on the effect of fin’s span-wise

stiffness on the propulsion performance was carried

out. The experiment investigated the thrust and lateral

force generation of five fin designs embedded with

different length of aluminum rods. The experimental

result shows that embedding a stiff rod into the fin’s

leading edge can cause the thrust and lateral force to

improve significantly at stroke frequency beyond

0.5Hz; up to 138% increasement in thrust for the HS-

1 (pink) relative to HS-0. In addition, the result also

suggests that the bionic fin named HS-0.5 (blue) takes

advantage over other designs in capacity of achieving

stable and fast motion for the prototype.

The fast motion performed by the prototype

shows the passively-flexing pectoral fin proposed in

our research is an excellent candidate for underwater

propulsive mechanism.

ACKNOWLEDGEMENTS

This research was financially supported by the

National Nature Science Foundation of China (No.

91748123).

REFERENCES

Yu, J., Liu, J., Wu, Z., & Fang, H.. (2017). Depth control of

a bioinspired robotic dolphin based on sliding mode

fuzzy control method. IEEE Transactions on Industrial

Electronics, 1-1.

Liu, H., Taylor, B., & Curet, O. M.. (2017). Fin ray stiffness

and fin morphology control ribbon-fin-based

propulsion. Soft Robot, 103-116.

Rosenberger, L. J.. (2001). Pectoral fin locomotion in

batoid fishes: undulation versus oscillation. Journal of

Experimental Biology, 204(Pt 2), 379-394.

Takemura, R., Akiyama, Y., Hoshino, T., & Morishima, K..

(2011). Chemical switching of jellyfish-shaped micro

robot consisting only of cardiomyocyte gel. Solid-state

Sensors, Actuators & Microsystems Conference. IEEE.

Breder, C. M. (1926). The locomotion of fishes.

Zoologica,4.

Fish, F. E., Kolpas, A., Crossett, A., Dudas, M. A., & Bart-

Smith, H.. (2018). Kinematics of swimming of the

manta ray: three-dimensional analysis of open-water

maneuverability. Journal of Experimental Biology,

221(Pt 6), jeb.166041.

Low, K. H., Zhou, C., Seet, G., Bi, S., & Cai, Y.. (2012).

Improvement and testing of a robotic manta ray

(RoMan-III). IEEE International Conference on

Robotics & Biomimetics. IEEE.

Cao, Y., Lu, Y., Cai, Y., Bi, S., & Pan, G.. (2019). Cpg-

fuzzy-based control of a cownose-ray-like fish robot.

Industrial Robot, ahead-of-print(ahead-of-print).

Liu, G., Ren, Y., Zhu, J., Bart-Smith, H., & Dong, H..

(2015). Thrust producing mechanisms in ray-inspired

underwater vehicle propulsion. Theoretical and

Applied Mechanics Letters, 5(001), 54-57.

Fish, F. E., Kolpas, A., Crossett, A., Dudas, M. A., & Bart-

Smith, H.. (2018). Kinematics of swimming of the

manta ray: three-dimensional analysis of open-water

maneuverability. Journal of Experimental Biology,

221(Pt 6), jeb.166041.

Parson, J. M., Fish, F. E., & Nicastro, A. J.. (2011). Turning

performance of batoids: limitations of a rigid body.

Design and Locomotion Control of a Myliobatid-inspired Robot Actuated by Passively-ﬂexing Pectoral Fins

359

Journal of Experimental Marine Biology & Ecology,

402(1-2), 12-18.

Fish, F.E., Schreiber, C.M., Moored, K.W., Liu, G., Dong,

H.and Bart Smith, H. (2016), “Hydrodynamic

performance of aquatic flapping: efficiency of

underwater flight in the manta”, Aerospace, Vol. 3 No.

3, pp. 1-24.

Clark, R. P., & Smits, A. J., (2006). Thrust production and

wake structure of a batoid-inspired oscillating fin.

Journal of Fluid Mechanics, 562(562), 415-429.

Dewey, P. A., Carriou, A., & Smits, A. J., (2012). On the

relationship between efficiency and wake structure of a

batoid-inspired oscillating fin. Journal of Fluid

Mechanics, 691, 245-266.

Dewey, P. A., Boschitsch, B. M., Moored, K. W., Stone, H.

A., & Smits, A. J., (2013). Scaling laws for the thrust

production of flexible pitching panels. JOURNAL OF

FLUID MECHANICS, 732, 29-46.

Chew, C. M., Lim, Q. Y., & Yeo, K. S. (2016).

Development of propulsion mechanism for Robot

Manta Ray. IEEE International Conference on Robotics

& Biomimetics. IEEE.

Wang, W., Gu, D., & Xie, G. (2017). Autonomous

optimization of swimming gait in a fish robot with

multiple onboard sensors. IEEE Transactions on

Systems Man & Cybernetics Systems, 1-13.

Zhou, C.L. and Low, K.H. (2012). Design and locomotion

control of a biomimetic underwater vehicle with fin

propulsion. IEEE/ASME Transactions on

Mechatronics, Vol. 17 No. 1, pp. 25-35.

Zhou, C.L. and Low, K.H. (2010a). Better endurance and

load capacity: an improved design of manta ray robot

(RoManII). Journal of Bionic Engineering, Vol. 7 No.

S4, pp. S137-S144.

Matsuoka, K. (1985). Sustained oscillations generated by

mutually inhibiting neurons with adaptation. Biological

Cybernetics, Vol. 52 No. 6, pp. 367-376.

Ijspeert, A.J., Crespi, A., Ryczko, D. and Cabelguen, J.-M.

(2007). From swimming to walking with a salamander

robot driven by a spinal cord model. Science, Vol. 315

No. 5817, pp. 1416-1420.

ICINCO 2022 - 19th International Conference on Informatics in Control, Automation and Robotics

360