Article

Development of Intelligent Drone Battery Charging

System Based on Wireless Power Transmission Using
Hill Climbing Algorithm
Ali Rohan 1, * , Mohammed Rabah 1 , Muhammad Talha 1 and Sung-Ho Kim 2
1 Department of Electrical, Electronics and Information Engineering, Kunsan National University,
Gunsan-Si 573-360, Korea; mohamedmostafamousa1991@gmail.com (M.R.); engrtalha72@gmail.com (M.T.)
2 Department of Control and Robotics Engineering, Kunsan National University, Gunsan-Si 573-360, Korea;
shkim@kunsan.ac.kr
* Correspondence: ali_rohan2003@hotmail.com; Tel.: +82-10-2857-6080




Received: 13 September 2018; Accepted: 5 November 2018; Published: 7 November 2018


Abstract: In this work, an advanced drone battery charging system is developed. The system is
composed of a drone charging station with multiple power transmitters and a receiver to charge the
battery of a drone. A resonance inductive coupling-based wireless power transmission technique
is used. With limits of wireless power transmission in inductive coupling, it is necessary that the
coupling between a transmitter and receiver be strong for efficient power transmission; however,
for a drone, it is normally hard to land it properly on a charging station or a charging device to get
maximum coupling for efficient wireless power transmission. Normally, some physical sensors such
as ultrasonic sensors and infrared sensors are used to align the transmitter and receiver for proper
coupling and wireless power transmission; however, in this system, a novel method based on the hill
climbing algorithm is proposed to control the coupling between the transmitter and a receiver without
using any physical sensor. The feasibility of the proposed algorithm was checked using MATLAB.
A practical test bench was developed for the system and several experiments were conducted under
different scenarios. The system is fully automatic and gives 98.8% accuracy (achieved under different
test scenarios) for mitigating the poor landing effect. Also, the efficiency η of 85% is achieved for
wireless power transmission. The test results show that the proposed drone battery charging system
is efficient enough to mitigate the coupling effect caused by the poor landing of the drone, with the
possibility to land freely on the charging station without the worry of power transmission loss.

Keywords: wireless power transfer; unmanned aerial vehicle; automatic charging station; drone
station; hill climbing

1. Introduction

1.1. Introduction and Motivation

The quadcopter, also known as a quadrotor, is a type of unmanned aerial vehicle (UAV), lifted
and propelled by four rotors [2,3]. Quadcopters use two pairs of identical fixed pitched propellers:
two clockwise and two counter-clockwise. The quadcopter has high maneuverability, as it can hover,
take off, cruise, and land in narrow areas. It also has a simpler control mechanism compared to other
UAVs [4] and is equipped with different components such as an inertial measurement unit (IMU),
a global positioning system (GPS), an electronic speed control (ESC), a standard radio control (RC),
a radio-frequency module (RF) used to transmit live videos to a personal computer (PC), and a flight
controller [5,6].

Appl. Syst. Innov. 2018, 1, 44; doi:10.3390/asi1040044 www.mdpi.com/journal/asi

Appl. Syst. Innov. 2018, 1, 44 2 of 19

The quadcopter is used for applications such as surveillance, search and rescue, and object
detection [7–9]. As mentioned earlier, a quadcopter with two pairs of fixed pitched propellers has
a very short operation time because it has to generate lift force all the time to move around, which
requires high electrical power. Batteries are used as an electrical power source in quadcopters; however,
because of the high electrical power requirement, the normal operation time of a quadcopter is
just 20 to 30 min [10]. This limits the quadcopter flight range and operation time drastically, and
accordingly, the quadcopter might not be able to fulfill the purpose of its use in a specific application.
Generally, to continuously operate quadcopters, the batteries are changed or recharged after the normal
operation time.
These batteries can be recharged using wired power transmission which requires some physical
connection or via wireless power transmission which does not require any physical connection.
Even though wired power transmission is more efficient than wireless power transmission, the wireless
power transmission technique is currently utilized for its lower maintenance and increased safety (due
to no physical connection of wires) for the delivery of power [11,12]. In order to charge the battery
using wireless power transmission, the quadcopter is equipped with electromagnetic coils. These coils
can be single or multiple depending upon the size and design of the charging system. Generally, a
wireless power transmission system is composed of a transmitting and a receiving side. Both the
transmitting and receiving sides are equipped with coils to transfer the power from the source to load.
The battery with the receiving coil is installed on the quadcopter and the transmitting side is normally
a ground station composed of a transmitting coil. For efficient power transmission, it is necessary that
the quadcopter land on the ground station in such a way that the receiving and transmitting coils are
aligned properly; however, due to the poor landing effect of quadcopters, there is always a chance of
misalignment. This misalignment causes power loss and affects the efficiency of the charging system.
To eliminate this misalignment issue, there is a need to develop a system which can easily mitigate the
poor landing effect. For that, a charging system is proposed in this work which can cope with such
issues and increase the power transmission efficiency.

1.2. Related Works

Previously, there were different researches on wireless power transmission. In References [13,14],
continuous inductive coupling in radio-frequency identification (RFID) tags were implemented.
In References [15,16], a wireless power technique based on an inductive coupling mechanism was
used for medical implants. Few wireless power systems for robotic applications were developed
using a large array of smaller coils driven by microelectromechanical systems (MEMS) switches and
organic field effect transistors to selectively transmit wireless power [17]. Some researchers developed
efficient wireless power transmission systems by designing a switched-mode direct current/alternating
current (DC/AC) inverter based on Class E topology [17–21]. Recently, there were many researches
on power control in wireless power transmission. In References [22–25], the authors proposed a
microcontroller-based power control method. In Reference [26], the authors developed a power control
scheme using analog feedback circuits. Some researchers used the closed-loop power control technique
using a commercial off-the-shelf (COTS) chipset [27–29].
For wireless charging of drones, some authors proposed laser beam systems which deliver the
power directly to the drone [30]. Solar energy to support a drone’s long flight time was proposed in
Reference [31]. Charging docking stations to recharge the drone battery were proposed in Reference [32].
Some authors applied smart contact arrays [33], whereas some proposed a drone charging station [34].
In Reference [35], the authors presented an approach based on optimal designing of the transmitting
and receiving coil with the goal of becoming less sensitive to the misalignment of coils. The system
was composed of a charging station with multiple arrays of primary or transmitting coils with a
specifically designed secondary or receiving coil. The receiving coil was designed to perfectly fit
in the landing skid of the drone. The authors proposed a transmitting coil overlapping scheme to
entirely cover the charging area on the charging station. By calculating the impedance of the multiple
Appl. Syst. Innov. 2018, 1, 44 3 of 19

transmitting coils and choosing the transmitting coil with maximum impedance, power transmission
between the transmitting and receiving coil was achieved. In Reference [36], the authors presented
a target detection technique based on image processing. After landing, the center of the coil was
aligned with the transmitting coil using a specific color detection and image-processing scheme.
The image-processing algorithm worked by taking the images using a drone camera and converted
red/green.blue (RGB) color space to hue/saturation/value (HSV) color space. After applying some
filters, the red color was detected and considered as a target. In References [37], the authors presented
a positioning system using a binary distance laser sensor and ultrasonic sensors. The system was
composed of a charging station comprising a single transmitting coil and a drone equipped with a
single receiving coil. The system worked by detecting the position of the receiving coil and aligning
the transmitting coil with it for wireless power transmission. The positioning system took almost 5 s to
detect and align with the receiving coil. Using sensors, the system complexity increased; it required
specific places and areas for installation on the drone and the charging station.

1.3. Contribution of the Paper and Road Map

To solve the problem of misalignment of coils caused by the drone’s imperfect landing, a battery
charging system based on wireless power transmission was developed in this work. The system is
composed of a ground charging station and a receiving coil with the load. The charging station is
equipped with multiple transmitting coils, whereas, on the receiver side, just one receiving coil is
used. Multiple transmitting coils are placed on a movable bed which can move in four directions
(positive X-direction, negative X-direction, positive Y-direction, and negative Y-direction). A control
technique based on the hill climbing algorithm was implemented to control the alignment between
the transmitting and receiving coil. The system was developed in a way allowing the drone to land
freely on the charging station without the worry of alignment between the receiving and transmitting
coil. The drone can land freely on charging station and the coils will adjust automatically in proper
alignment to start the power transmission. Previously, the misalignment issues were solved using
different methods, such as the design of a charging station with overlapped coils [35], using a target
detection technique based on image processing [36], and using a positioning system based on physical
sensors [37]. The proposed system is based on a novel method which uses the hill climbing algorithm
on backscattered voltage signals of the transmitting coils. It improves the accuracy of the system and
eliminates the flaws found in previous techniques such as the image-processing technique, where
there are chances of missing the target. Also, it gives the freedom for the drone to land freely on
a charging station and makes the system more adoptable by avoiding the use of physical sensors.
The configuration of the proposed system is discussed in detail in Section 2.

2. Configuration of the Proposed Wireless Battery Charging System for a Quadcopter

Figure 1 shows the proposed wireless battery charging system for a quadcopter. The system is
composed of the following three parts:

1. Wireless power transmitter, comprising transmitting coil array which can move in four directions.
2. Wireless power receiver, a receiving coil, and battery charger.
3. Control unit which can measure the terminal voltage of each transmitting coil, and align the
centroid of the transmitting and receiving coil using the hill climbing algorithm.
Appl. Syst. Innov. 2018, 1, 44 4 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 4 of 19

Figure 1.
Figure Blockdiagram
1. Block diagramof
ofthe
the proposed
proposed wireless
wireless battery
battery charging
charging system
system for
for aa quadcopter.
quadcopter.

The proposed wireless battery charging system consists of multiple transmitting coils which
The proposed wireless battery charging system consists of multiple transmitting coils which can
can be moved in four directions (positive X-direction, negative X-direction, positive Y-direction, and
be moved in four directions (positive X-direction, negative X-direction, positive Y-direction, and
negative Y-direction). The use of multiple coils can potentially allow the system to efficiently adapt
negative Y-direction). The use of multiple coils can potentially allow the system to efficiently adapt
to magnetic field propagation conditions, similar to the way multiple antennas are used to adapt to
to magnetic field propagation conditions, similar to the way multiple antennas are used to adapt to
channel conditions in wireless communication systems [38]. Also, multiple transmitting coils decrease
channel conditions in wireless communication systems [38]. Also, multiple transmitting coils
the time for coil alignment, providing a chance for the design of big charging stations for large drones,
decrease the time for coil alignment, providing a chance for the design of big charging stations for
where coil size and design are limited due to power transmission characteristics. In this proposed
large drones, where coil size and design are limited due to power transmission characteristics. In this
system, if a drone with a receiving coil lands at any position on the multiple transmitting coils, the
proposed system, if a drone with a receiving coil lands at any position on the multiple transmitting
voltages across the transmitting coils decrease depending on the coupling of each transmitting coil
coils, the voltages across the transmitting coils decrease depending on the coupling of each
with the receiver coil, and the controller knows where the drone lands by observing the change in
transmitting coil with the receiver coil, and the controller knows where the drone lands by observing
terminal voltages of the multiple coils. When the power is transferred from the transmitting coil to the
the change in terminal voltages of the multiple coils. When the power is transferred from the
receiving coil, the terminal voltage of the transmitting coil tends to decrease due to the phenomenon
transmitting coil to the receiving coil, the terminal voltage of the transmitting coil tends to decrease
called signal backscattering.
due to the phenomenon calledBy observing
signal the backscattered
backscattering. signal
By observing thefrom each transmitting
backscattered signal fromcoil,each
the
controller automatically
transmitting knows which
coil, the controller transmitting
automatically coil is which
knows nearest transmitting
to the receivingcoilcoil.
is If the controller
nearest to the
pinpoints the nearest transmitting coil, the multiple transmitting coil array is moved to automatically
receiving coil. If the controller pinpoints the nearest transmitting coil, the multiple transmitting coil
align the
array centroid
is moved toof the detected transmitting
automatically coil with
align the centroid thatdetected
of the of the receiving coil using
transmitting coil the
withhill climbing
that of the
algorithm. The practical system used to implement the proposed system comprises a
receiving coil using the hill climbing algorithm. The practical system used to implement the proposed receiver circuit
for battery
system recharging
comprises and a circuit
a receiver power for
transmission station. and a power transmission station.
battery recharging
2.1. Transmitter and Receiver Circuit Design and Description
2.1. Transmitter and Receiver Circuit Design and Description
Figure 2 shows a simple wireless power transmission circuit. The circuit comprises a transmitter
Figure 2 shows a simple wireless power transmission circuit. The circuit comprises a transmitter
circuit, a receiving circuit, and the control part (voltage divider and analog-to-digital converter (ADC)
circuit, a receiving circuit, and the control part (voltage divider and analog-to-digital converter (ADC)
of the microcontroller). The power amplifier drives the transmitting coil Lt . When the receiving coil
of the microcontroller). The power amplifier drives the transmitting coil 𝐿 . When the receiving coil
is brought near to the transmitting coil, the voltage at receiving coil Lr is induced and it is processed
is brought near to the transmitting coil, the voltage at receiving coil 𝐿 is induced and it is processed
by a bridge rectifier to convert AC to DC voltage. To achieve the optimum performance, values of
by a bridge rectifier to convert AC to DC voltage. To achieve the optimum performance, values of
Ct , Lo , Co , and Cr are calculated using Equations (1) to (4).
𝐶 , 𝐿 , 𝐶 , and 𝐶 are calculated using Equations (1) to (4).
Appl. Syst. Innov. 2018, 1, 44 5 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 5 of 19

Figure 2. Basic
Figure 2. Basic circuit
circuit diagram
diagram for
for wireless
wireless power
power transmission.
transmission.

11
R𝑅0 L𝐿r ±
± 2 ωM𝜔𝑀2
C𝐶r == 2 ;; (1)
(1)
R𝑅0 ω 2 2
𝜔 L𝐿r
−1
C𝐶0 == 𝜔
ω
; (2)
− 2 ) + R ( Q + 1 − sec( ϕ )) ; (2)
𝜔𝐿t (1 − 𝐾 ) + 𝑅0 (𝑄 + −
ωL ( 1 K (𝜑))
2ω −1 )
(2𝜔


C𝐶t == ;; (3)
(3)
11 + 𝜋
π2
+ 4 R𝑅
4
L0 = ω −1 QR0 . (4)
𝐿 = 𝜔 𝑄𝑅 . (4)
In
In the
the above equations, M
above equations, 𝑀 isis the
the mutual inductance, K
mutual inductance, 𝐾 is
is the
the coupling factor, Q
coupling factor, 𝑄 isis the
the quality
quality
factor between the transmitting and receiving coils, and is the phase angle ranging from 40 ◦ to 70◦ .
factor between the transmitting and receiving coils, and 𝜑 is the phase angle ranging from 40° to 70° .
ϕ
The
The derivation
derivation of of the
the equations
equations for for aa Class-E
Class-E amplifier
amplifier cancan bebe found
found in in Reference
Reference [39].[39]. Transmitting
Transmitting
coils
coils and
and thethe receiving
receivingcoil coilare
aremade
madeofofcopper
copper wire
wire with
with 15 15andand 40 turns,
40 turns, respectively
respectively (Tables(Tables
1 and1
and 2). According to the power transmission range capabilities, wireless
2). According to the power transmission range capabilities, wireless power transmission can be power transmission can be
categorized
categorized into into three
three types
types [40,41].
[40,41]. First
First is
is the
the inductive
inductive powerpower transmission
transmission (IPT) (IPT) andand capacitive
capacitive
power
power transmission (CPT), used for short-range distances; second is the resonant inductive coupling
transmission (CPT), used for short-range distances; second is the resonant inductive coupling
power
powertransmission,
transmission, widely
widelyused for medium-range
used for medium-range distances; and third
distances; and is the laser
third is beam or microwave
the laser beam or
power
microwavetransmission, used for long-range
power transmission, used for distances.
long-range In order toIn
distances. achieve
order tohigh efficiency,
achieve inductive
high efficiency,
coupling
inductiverequires
couplingvery requiresclose coupling
very betweenbetween
close coupling the transmitting and receiving
the transmitting and receivingcoil. coil.
Whereas,
Whereas,in
resonant inductive coupling, efficient power transmission can be achieved
in resonant inductive coupling, efficient power transmission can be achieved with some distance with some distance between
the transmitting
between and receiving
the transmitting coil via the
and receiving use
coil via ofthe
resonant
use of circuits.
resonantAlso, resonant
circuits. Also, inductive coupling
resonant inductive
has better tolerance than inductive coupling. Therefore, the resonant inductive
coupling has better tolerance than inductive coupling. Therefore, the resonant inductive coupling is coupling is considered
an effective an
considered technique
effectivefor coping with
technique the coil
for coping withmisalignment issues, andissues,
the coil misalignment for drone
and battery
for drone charging
battery
systems. Therefore, in this work, a resonance inductive coupling-based
charging systems. Therefore, in this work, a resonance inductive coupling-based wireless power wireless power transmission
technique
transmission wastechnique
used for was charging
used forthe charging
drone battery. Thebattery.
the drone detailedThe block diagram
detailed block of diagram
the circuit of for
the
wireless power transmission is shown in Figure 3. At the transmitter side,
circuit for wireless power transmission is shown in Figure 3. At the transmitter side, Class-E power Class-E power amplifiers
are used to are
amplifiers generate
used amplified
to generate AC voltages AC
amplified withvoltages
a resonancewithfrequency
a resonance of 240 kHz forof
frequency each
240coil.
kHz for
each coil.
Appl. Syst. Innov.
Appl. Syst. Innov. 2018, 1, x
2018, 2, 44FOR PEER REVIEW 66of
of 19
19

Figure 3.
Figure Block diagram
3. Block diagram of
of the
the circuit
circuit for
for wireless
wireless power
power transmission
transmission with
with multiple
multiple transmitting
transmitting coils
coils
and a receiving coil.
and a receiving coil.
Table 1. Excitation circuit component values.
Table 1. Excitation circuit component values.
Component Value
Component Value
Number of Turns 15
Number of Turns 15
Lt 𝐿 14.5µH
14.5 µH
Lc 1 mH
𝐿 1 mH
Lo 9.9 µH
𝐿 9.9 µH
Ct 15 nF
Co
𝐶 15 nF
28 nF
R1 𝐶 28 nF
100 Ω
R2 𝑅 100 Ω
30 Ω
𝑅 30 Ω
Table 2. Receiver circuit component values.

Table Component Valuevalues.

2. Receiver circuit component
Number of Turns 40
Component Value
L r
Number of Turns 733.39
40 µH
Cr 680 pf
𝐿 733.39 µH
CL 0.05 µF
RLoad 𝐶 6802 kΩpf
𝐶 0.05 µF
𝑅 2 kΩ
These amplifiers are driven by a high-frequency clock signal. The clock signal is provided to the
gate driver of each amplifier of the respective transmitting coil simultaneously. A microcontroller is
used for generating a high-frequency clock signal and it keeps measuring the output voltages of each
transmitting coil using an ADC, and performs envelop detection of the measured voltages to identify
which excitation coil is the closest to the the receiving
receiving coilcoil of
of the
the drone.
drone.
At the
the receiving
receivingside,
side,the
thereceiving
receiving coilcoil
is installed on the
is installed on drone and itand
the drone is connected to a full
it is connected tobridge
a full
rectifier,
bridge which which
rectifier, is usedisto convert
used the AC
to convert thetoAC DC. Finally,
to DC. a battery
Finally, charger
a battery charger(DC(DC to DC converter)
to DC converter)is
used for battery charging. The system works by detecting the change
is used for battery charging. The system works by detecting the change in the voltage in the voltage of any of the
envelope-detected voltage
envelope-detected voltage signals
signals ofof the transmitting
transmitting coils,
coils, which
which (changes
(changes in voltage) refer to the
presence of the receiving coil on on the
the charging
charging station.
station. After detecting the receiving coil, a control
algorithm is activated to align the coils and start wireless power power transmission.
transmission.

2.2. Four-Way
2.2. Four-Way Directional
Directional XY Table
XY Table
A four-way
A directional XY
four-way directional table was
XY table was constructed to control
constructed to control the
the position
position of
of the
the transmitting coil
transmitting coil
array, as shown in Figure 4. The XY table is controlled by two stepper motors that are driven
array, as shown in Figure 4. The XY table is controlled by two stepper motors that are driven by twoby two
stepper motor
stepper motor drivers. Motor 11 is
drivers. Motor is responsible
responsible for
for moving
moving the
the transmitting
transmitting coil
coil array
array in
in the Y-direction,
the Y-direction,
while motor 2 is responsible for moving it in the X-direction. When the quadcopter approaches the
Appl. Syst. Innov. 2018, 1, 44 7 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 7 of 19

while motor
charging 2 is responsible
station, for moving
the transmitting coil, already X-direction.
it in theexcited havingWhen the quadcopter
a constant approachesa
voltage, experiences
the charging station, the transmitting coil, already excited having a constant voltage, experiences
change in voltage. This change in voltage refers to the presence of the receiving coil and load (battery). a
change in voltage.
By measuring theThis change
voltage in voltage
change from refers to the presence
each transmitting of the
coil, the receiving
charging coil and is
station load (battery).
moved in a
By measuring
direction where theone
voltage
of the change from coils
transmitting each istransmitting coil,
aligned to the the charging
receiving station
coil. The is moved
controller sendsinthe
a
direction where one of the transmitting coils is aligned to the receiving coil. The
required pulse width modulation (PWM) signals to the stepper motor driver to move the transmittingcontroller sends the
required
coil arraypulse width
to the modulation
proper position to(PWM) signals
initiate to the stepper
the wireless power motor driver to
transmission move
and the transmitting
battery charging.
coil array to the proper position to initiate the wireless power transmission and battery charging.

Figure 4. Four-way directional XY table.

Figure 4. Four-way directional XY table.
2.3. Control Part
2.3. Control Part
The control part of the proposed system performs specific tasks such as generating the clock signal
The control
for driving part ofcoils,
transmitting the proposed
monitoring system performs
each coil’s specific
terminal tasksand
voltage, suchcontrolling
as generating the clock
the four-way
signal for driving transmitting coils, monitoring each coil’s
XY table to align the centroids of the transmitting and receiving coils. terminal voltage, and controlling the four-
wayGenerally,
XY table to it align
is notthe
easycentroids
to land aofdrone
the transmitting
at a specificand
point receiving coils. station precisely. In most
on the drone
Generally, it is not easy to land a drone at a specific point on
cases, the centroids of the transmitting and receiving coil are misaligned, as shown the drone station inprecisely.
Figure 5.In most
cases, the centroids
When of the transmitting
the misalignment betweenand thereceiving coil are the
coils happens, misaligned,
efficiencyas shown
of the in Figure power
wireless 5.
When the misalignment between the coils happens, the efficiency of
transmission deteriorates. In order to solve this problem, an intelligent automatic alignment algorithm the wireless power
transmission deteriorates. In order to solve this
based on the hill climbing algorithm is proposed and implemented. problem, an intelligent automatic alignment
algorithm based onalignment
The automatic the hill climbing
algorithm algorithm is proposed
was implemented andaimplemented.
inside microcontroller. The microcontroller
The automatic alignment algorithm was implemented
keeps measuring the terminal voltages of each transmitting coil simultaneously, inside a microcontroller.
and finds out which The
microcontroller keeps measuring the terminal voltages of each transmitting
transmitting coil has the lowest voltage. Generally, when the receiving coil of the drone is near tocoil simultaneously, and
afinds outtransmitting
certain which transmitting
coil, the coil has the lowest
corresponding voltage. Generally,
transmitting when
coil’s terminal the receiving
voltage tends tocoil of the
decrease.
drone is near to a certain transmitting coil, the corresponding transmitting coil’s
Therefore, the microcontroller can detect the transmitting coil with a voltage difference and, from this terminal voltage
tends toon,
moment decrease. Therefore, the
the microcontroller microcontroller
tries can detect
to align the centroid of thethe transmitting
transmitting coilcoil
withwith a voltage
the receiving
difference and, from this moment on,
coil by moving the four-way directional XY table.the microcontroller tries to align the centroid of the transmitting
coil with the receiving coil by moving the four-way directional XY table.
 Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 8 of 19
Appl. Syst. Innov. 2018, 1, 44 8 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 8 of 19

Figure 5.
Figure 5. Misalignment of coils
coils for
for wireless
wireless power
power transmission.
transmission.

2.3.1. Figure 5. Misalignment of coils for wireless power transmission.

2.3.1. Signal
Signal Backscattering
Backscattering
2.3.1. Signal
Signal backscattering
backscattering is
Backscattering is aa phenomenon
phenomenon basicallybasically used used in in wireless
wireless communication
communication for for RFID
RFID
(radio-frequency
(radio-frequency identification).
identification). The transmitting
The transmitting side voltage
side used
voltage tends to decrease by increasing the
Signal backscattering is a phenomenon basically in tends
wireless to decrease
communicationby increasing
for RFID the
coupling
coupling between the
between thetransmitting
transmittingand and receiving
receiving coils.
coils. When Whenthe the drone
drone lands lands
on the on the charging
charging station,
(radio-frequency identification). The transmitting side voltage tends to decrease by increasing the
station,
the XY the XYstarts
table tablethestarts the aligning
aligning process process
and, coils.and, during
during the the
time theoftime of movement, the voltage
coupling between the transmitting and receiving When dronemovement,
lands on thethe voltage
charging at the
station,
at the transmitting
transmitting side side
decreasesdecreases
rapidly. rapidly.
This This
change change
in in
voltage voltage
produces produces
a naturala natural backscattered
backscattered signal.
the XY table starts the aligning process and, during the time of movement, the voltage at the
signal. This backscattered
This backscattered signal at signal at ahigh
a very very high sampling frequency of 84isMHz is continuously byread
transmitting side decreases rapidly. This sampling
change in frequency
voltage producesof 84 MHz a naturalcontinuously
backscattered readsignal.the
by the ADC
ADCbackscattered of the microcontroller.
of the microcontroller. Figure 6 shows the sampling process of the voltage signal during
This signal at a Figure
very high6 shows
samplingthe sampling
frequencyprocessof 84 MHz of the is voltage
continuouslysignalreadduring one
by the
one
periodperiod
of of time
time for for aligned
aligned (transmitting
(transmitting coil coil voltage
voltage at load)atand
load) and misaligned
misaligned (transmitting
(transmitting coil coil
voltage
ADC of the microcontroller. Figure 6 shows the sampling process of the voltage signal during one
voltage
at no load) at no load)
coils. coils.ofInside
Inside of the microcontroller,
the microcontroller, an envelope an envelope
detection detection
algorithm algorithm
is used to is usedthe
detect to
period of time for aligned (transmitting coil voltage at load) and misaligned (transmitting coil voltage
detect
envelope the envelope of the backscattered signal. After the detection of an envelope, the peak value of
at no load)ofcoils.
the backscattered signal. After the an
Inside of the microcontroller, detection
envelope of an envelope,
detection the peakisvalue
algorithm used of to the voltage
detect the
the voltage
signal is of signal isand
selected selected and this
this peak valuepeak value is observed
is the
observed continuouslycontinuously
to provide to provide
thevalue the information
information about
envelope the backscattered signal. After detection of an envelope, the peak of the voltage
about
aligned aligned or misaligned coils. In the case of aligned coils, the peak value will be very low (almost
signal is or misaligned
selected coils.peak
and this In the case is
value of observed
aligned coils, the peak value
continuously will bethe
to provide very low (almostabout
information 25 V;
25 V; Figure 6) inand, incase
the ofcase of misalignment, the peak value will be high (almost 70 V; Figure 6).
aligned or misaligned coils. In the case of aligned coils, the peak value will be very low (almost 25The
Figure 6) and, the misalignment, the peak value will be high (almost 70 V; Figure 6). V;
The
hill hill climbing algorithm
climbing processes this voltage informationdata dataand andfinds
findsthe the optimum solutionsolution by
Figure 6) and, algorithm
in the caseprocesses this
of misalignment, voltagethe information
peak value will be high (almostoptimum
70 V; Figure 6). The by
moving
moving the XY
the XY table in a specific
table in processes direction
a specific direction and aligning
andinformation the
aligning thedata transmitting
transmitting and receiving coils.
hill climbing algorithm this voltage and findsand the receiving
optimumcoils. solution by
moving the XY table in a specific direction and aligning the transmitting and receiving coils.

Figure 6.
Figure 6. Sampling
Sampling of
of voltage
voltage signal
signal during
during one
one time
time period.
period.

Figure 6. Sampling of voltage signal during one time period.

Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 9 of 19
Appl. Syst. Innov. 2018, 1, 44 9 of 19
2.3.2. Hill Climbing Algorithm
Hill climbing
2.3.2. Hill Climbing is anAlgorithm
optimization technique that is used to find an optimum solution to a
computational problem. It starts off with a solution that is normally very poor compared to the
Hill climbing
optimal solution and then is an optimization
iteratively improves technique
from there.thatItisdoes
usedthis
to by
find an optimum
generating other solution
solutionsto a
computational problem. It starts off with a solution that is normally
which are better than the current solution. It repeats the process until it finds the optimal solution very poor compared to the
whereoptimal
it can solution
no longer and then
find anyiteratively
improvements. improves from there. It does this by generating other solutions
which
Generally, drones tend to land at any placeItupon
are better than the current solution. repeats
the the process
drone’s until battery
wireless it finds charging
the optimal solution
station.
where it can no longer find any improvements.
That means the centroid of the receiving coil on the drone may not be aligned with any of the
transmittingGenerally,
coils. Indrones tend to
this case, land at anycoils
transmitting placeare upon the drone’s
moved using thewireless
XY table batteryin charging
an arbitrary station.
That means the centroid of the receiving coil on the drone may
direction to get closer to the receiving coil. By measuring terminal voltages of each transmitting coil, not be aligned with any of the
transmitting coils. In this case, transmitting coils are moved using the
the controller can detect whether the receiving coil gets closer to one of the transmitting coils. This XY table in an arbitrary direction
kindtoofget closermovement
arbitrary to the receiving of the XY coil. Bycontinues
table measuring terminal
until voltages
it detects of each
the decrease in transmitting
terminal voltages coil, the
of the transmitting coils, as shown in Figure 7. When one transmitting coil, which correspondskind
controller can detect whether the receiving coil gets closer to one of the transmitting coils. This to of
arbitrary coil
transmitting movement
2 in Figureof the 7,XYistable chosen,continues
the hill until it detects
climbing the decrease
algorithm in terminal
is activated to voltages
move the of the
transmitting coils, as shown in Figure 7. When one transmitting
transmitting coil to the proper position where the voltage measured from that transmitting coil coil, which corresponds to transmitting
coil its
reaches 2 inminimum.
Figure 7, is chosen, the hill climbing algorithm is activated to move the transmitting coil to the
proper
Figureposition
8 showswhere the voltage
the overall flowchart measured
of the from
proposedthat transmitting
method. Figure coil9reaches
shows its theminimum.
flowchart of
Figure 8 shows the overall flowchart of the proposed method.
the hill climbing algorithm used in this system. Previously, authors tried to solve a different Figure 9 shows the flowchart
kinds of of
the hill climbing algorithm used in this system. Previously, authors
control problems using the hill climbing algorithm [42,43]. In this work, the hill climbing algorithm tried to solve a different kinds of
control
starts problems
by sending using thePWM
the required hill climbing
value toalgorithm
the stepper [42,43].
motorIn this work,
driver to move thethe hillXY
climbing
table. At algorithm
the
starts by sending the required PWM value to the stepper motor driver
start, when there is no change in the voltage, i.e., the drone is not on the charging station, there is no to move the XY table. At the
start, when
movement and thethere is no change
voltage in the voltage,
of the transmitting coili.e., the droneWhen
is constant. is notthe
ondrone
the charging
lands on station, there is no
the charging
movement and the voltage of the transmitting coil is constant. When
station, there is a change in the voltage value, and the voltage decreases in the presence of a drone the drone lands on the charging
with a receiving coil. The hill climbing algorithm is activated and one transmitting coil with athe
station, there is a change in the voltage value, and the voltage decreases in the presence of drone
with a change
maximum receiving coil.
in the The hill
voltage value climbing algorithm
is selected. is activated
The current terminal and one transmitting
voltage 𝑉 of the selected coil with
coil the
maximum change in the voltage value is selected.
is measured and compared with the previous measured voltage 𝑉 . The minimum value The current terminal voltage V n of the selected
is stored
as 𝑉 . After that, the controller sends the required PWM to move the XY table to the positionvalue
coil is measured and compared with the previous measured voltage V n − 1 . The minimum of is
stored as V min . After that, the controller sends the required PWM
the minimum value. This kind of process keeps repeating until it reaches the minimum voltage. In to move the XY table to the position
of the
Figure 9, 𝑛minimum
= 1 to 4 isvalue. This kind
the number of process keeps
of transmitting coils, repeating
and 𝑘 is the until it reaches
number of the the minimum voltage.
sample.
In Figure 9, n = 1 to 4 is the number of transmitting coils, and k is the number of the sample.

Figure 7. (a) Receiving coil’s initial position. (b) Receiving coil gets close to a transmitting coil after
Figure 7. (a) Receiving coil’s initial position. (b) Receiving coil gets close to a transmitting coil after
arbitrary movement.
arbitrary movement.
 Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 10 of 19
Appl. Syst. Innov. 2018, 1, 44 10 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 10 of 19

Figure 8. Overall flowchart of the proposed method.

Figure 8. Overall flowchart of the proposed method.
Figure 8. Overall flowchart of the proposed method.

Figure 9. Flowchart of the hill climbing algorithm.

Figure 9. Flowchart of the hill climbing algorithm.
Figure 9. Flowchart of the hill climbing algorithm.
Appl. Syst. Innov. 2018, 1, 44 11 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 11 of 19

3. Explanation
3. Explanation of
of the
the Test
Test Bench
Bench
In order
In order to
to verify
verify the
the feasibility of the
feasibility of the proposed
proposed algorithm,
algorithm, we we made
made aa test
test bench
bench of
of aa wireless
wireless
power transmission and battery charging station for the drone. The developed test bench is shown
power transmission and battery charging station for the drone. The developed test bench is shown
in Figure
in Figure 10.
10.The
Thetest
test bench
bench was
was composed
composed of four
of four transmitting
transmitting coilscoils
whichwhich are mounted
are mounted on theonfour-
the
four-way XY table, one receiving coil connected to an electrical load, and a controller
way XY table, one receiving coil connected to an electrical load, and a controller which performs thewhich performs
the measurement
measurement andand
hill hill climbing
climbing algorithm.
algorithm.

Figure 10. Practical

Practical test
test bench for drone battery charging station.

3.1. Battery
3.1. Battery Charging
Charging Station
Station
The battery
The battery charging
charging station
station waswas built
built using
using an an XYXY table.
table. The The XYXY table
table is
is used
used to to move
move the
the
transmitting coil
transmitting coil array
array to to the
the position
position where
where thethe centroid
centroid of of the
the transmitting
transmitting coil coil is
is aligned
aligned with
with the
the
centroid of the receiving coil using the hill climbing algorithm. The dimensions
centroid of the receiving coil using the hill climbing algorithm. The dimensions of the XY table were of the XY table were
394
394mmmm××414 414mm mm××93.6 93.6mm mm ×w
(l (𝑙 ×× ). The
𝑤 h× ℎ ). main
The main of the XY
frame frame of table
the XY wastable
madewas of aluminum
made of
aluminum and the rectangular plate, where the coils are placed, was made of a plastic sheet withof
and the rectangular plate, where the coils are placed, was made of a plastic sheet with a thickness a
2.5 mm. Four
thickness transmitting
of 2.5 coils were placed
mm. Four transmitting coils on
weretopplaced
of the on
rectangular
top of theplate within the
rectangular XYwithin
plate table and
the
theytable
XY wereandcontrolled
they were by controlled
two stepper bymotors for positioning.
two stepper motors for positioning.
In this work, four excitation circuits for the four transmitting
In this work, four excitation circuits for the four transmitting coils werecoilsdeveloped. Each transmitting
were developed. Each
coil was connected to the excitation circuit. A gate voltage of 15 V and
transmitting coil was connected to the excitation circuit. A gate voltage of 15 V and a supply a supply voltage of 12 voltage
V were
applied
of to theapplied
12 V were IRF510to metal-oxide-semiconductor field-effect transistor
the IRF510 metal-oxide-semiconductor field-effect(MOSFET),
transistor and it was driven
(MOSFET), and
by a low-power 240 kHz clock signal, generated by the microcontroller.
it was driven by a low-power 240 kHz clock signal, generated by the microcontroller. A voltage A voltage divider circuit was
also builtcircuit
divider to decrease
was alsothe built
output tovoltage
decrease into
thea output
voltage voltage
that wasintoreadable by the
a voltage controller.
that Based by
was readable on the
the
proposed method
controller. Based andon the the equations
proposed presented
method and in Reference [40], the
the equations optimumin
presented values of the electrical
Reference [40], the
optimum values of the electrical components for the excitation circuits were calculated.values
components for the excitation circuits were calculated. Table 1 shows the component Table 1used
showsin
the excitation
the component circuit,
values and Figure
used in 11
theshows a closer
excitation view and
circuit, of theFigure
excitation circuitsa and
11 shows controller.
closer view of the
excitation circuits and controller.
Appl. Syst. Innov. 2018, 1, 44 12 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 12 of 19

Figure 11. Closer

Closer view of the excitation circuits and controller.

3.2. Receiving Coil

3.2. Receiving Coil and
and Electric
Electric Load
Load
When
When the the drone
drone lands
lands on on the
the battery
battery charging
charging station
station and the XY
and the XY table
table starts
starts moving
moving the the
transmitting
transmitting coil coil array
arraytotothe
theproper
properposition
position using
using thethe
hillhill climbing
climbing algorithm,
algorithm, the the transmitting
transmitting coil
coil starts
starts to couple
to couple inductively
inductively with thewith the receiving
receiving coil.inductive
coil. This This inductive
coupling coupling
betweenbetween
the coilsthe coils
induces
induces voltage in the receiving coil. The receiving coil is connected to the receiving
voltage in the receiving coil. The receiving coil is connected to the receiving circuit as shown in Figure circuit as shown
in Figure
3. The 3. Thevoltage
induced inducedatvoltage at Lr is processed
𝐿 is processed by a full-wave
by a full-wave bridge rectifier
bridge rectifier to convert
to convert AC to DC. AC In
to this
DC.
In this system, rectification is achieved via four insulated-gate bipolar transistors
system, rectification is achieved via four insulated-gate bipolar transistors (IGBTs), designed to work (IGBTs), designed
to
onwork on high-frequency
high-frequency AC input AC input signals.
signals. The voltage
The output output voltage is obtained
is obtained across
across the loadthe load resistor
resistor 𝑅 .
R . Table 2 shows the component values used in the receiving circuit, and
Table 2 shows the component values used in the receiving circuit, and Figure 12 shows a closer view
Load Figure 12 shows a closer
view
of theofreceiving
the receiving
circuitcircuit and receiving
and receiving coil. coil.

Figure 12. Closer view of the receiving circuit and receiving

receiving coil.
coil.

3.3. Controller
3.3. Controller
In order to perform the control tasks, an STM32f4 discovery kit featuring a 32-bit ARM Cortex-M4
In order to perform the control tasks, an STM32f4 discovery kit featuring a 32-bit ARM Cortex-
was used. It can generate up to a 168-MHz clock signal. It also supports ADC with 19 channels and
M4 was used. It can generate up to a 168-MHz clock signal. It also supports ADC with 19 channels
12-bit resolution. The controller was used to generate 240-kHz clock signals for the four excitation
and 12-bit resolution. The controller was used to generate 240-kHz clock signals for the four excitation
circuits and it read the terminal voltage of transmitting coils simultaneously.
circuits and it read the terminal voltage of transmitting coils simultaneously.
When the drone lands on the battery charging station, the controller sends different PWM values
When the drone lands on the battery charging station, the controller sends different PWM values
to the two stepper motor drivers to move the XY table in random directions. At the same time, the
to the two stepper motor drivers to move the XY table in random directions. At the same time, the
controller
controller keeps
keeps measuring the output
measuring the voltage of
output voltage of the
the four
four transmitting coils. Once
transmitting coils. Once there
there is
is aa drop
drop in
in
the voltage across one of the transmitting coils, the controller triggers the hill climbing algorithm and
sends the required PWM to the stepper motor drivers until it measures the minimum voltage value
Appl. Syst. Innov. 2018, 1, 44 13 of 19

the
Appl.voltage across
Syst. Innov. 2018, one of the
2, x FOR transmitting
PEER REVIEW coils, the controller triggers the hill climbing algorithm and
13 of 19
sends the required PWM to the stepper motor drivers until it measures the minimum voltage value
transmitting and receiving coils are aligned; then, the hill climbing
where both the centroids of the transmitting
algorithm stops working, and the wireless power transmission begins until the the drone
drone is
is fully
fully charged.
charged.

4. Experiments and Results

Several experiments are conducted conducted with a test bench in order to verify the feasibility of the
proposed scheme. The The system
system waswas tested
tested under
under different test scenarios depending on the drone
landing position on the charging station. The receiving coil attached with load was placed at different
positions
positions on onthethecharging
chargingstation
stationand
andthetheresponse
response of of
thethesystem
system waswasobserved.
observed. In one
In onetesttest
scenario, the
scenario,
system was tested for the misalignment of x = 100 mm and y = 50 mm,
the system was tested for the misalignment of x = 100 mm and y = 50 mm, whereas in other test whereas in other test scenarios,
the misalignment
scenarios, was x = 75was
the misalignment mm,x y= =7530 mmyand
mm, = 30xmm = 50andmm, x =y 50
= 10mm,mm. y =Regardless
10 mm. Regardlessof the position
of the
of the receiving
position coil, the charging
of the receiving coil, thestation
charging wasstation
able towasperfectly
able to align the centroid
perfectly align theof the transmitting
centroid of the
and receivingand
transmitting coilreceiving
with an accuracy
coil with ofan98.8%.
accuracy Figure 13 shows
of 98.8%. Figureone13ofshows
the testonescenarios.
of the test Initially,
scenarios.we
assumed
Initially, wethatassumed
the receiving
that thecoil and fourcoil
receiving transmitting coils were positioned
and four transmitting coils were as shown inasFigure
positioned shown13a. in
At the start, the XY table was moved randomly until the receiving coil
Figure 13a. At the start, the XY table was moved randomly until the receiving coil got closer to onegot closer to one of the four
transmitting coils. During coils.
of the four transmitting this process,
Duringthe thismicrocontroller kept measuring the
process, the microcontroller keptfour terminal the
measuring voltages
four
simultaneously.
terminal voltages Bysimultaneously.
detecting the voltage drop between
By detecting the voltageall four
drop transmitting
between allcoils,fourthe microcontroller
transmitting coils,
activated the hill climbing
the microcontroller activatedalgorithm
the hill and the XY
climbing table was
algorithm andmoved
the XY to align
table the
wasnearest
movedtransmitting
to align the
coil withtransmitting
nearest a receiving coil coil.with
The distance
a receiving and position
coil. of the nearest
The distance transmitting
and position coil were
of the nearest judged on
transmitting
the
coil basis
were of the voltage
judged on the drop.
basis of The
thecloser
voltage thedrop.
transmitting
The closer andthereceiving coil, and
transmitting the larger
receiving the coil,
voltage
the
drop would be due to power transmission between the two coils. As
larger the voltage drop would be due to power transmission between the two coils. As shown in shown in Figure 13b,c, after the
random movement
Figures 13b,c, of the
after the XY table,
random transmitting
movement of the XYcoiltable,
1 was found as the
transmitting nearest
coil coil to as
1 was found thethereceiving
nearest
coil. Then, the hill climbing algorithm was activated by the microcontroller
coil to the receiving coil. Then, the hill climbing algorithm was activated by the microcontroller andand the XY table moved to
align
the XY transmitting
table moved coil
to 1align
withtransmitting
the receivingcoil coil. Figure
1 with the14receiving
shows the recorded
coil. Figure voltage
14 shows waveforms
the recorded for
the scenario
voltage shown infor
waveforms Figure 13. It canshown
the scenario be seeninclearly
Figurethat13.the voltage
It can of transmitting
be seen clearly thatcoil the1 voltage
decreased, of
red in the case of Figure 13a, green in the case of Figure 13b, and blue in
transmitting coil 1 decreased, red in the case of Figure 13a, green in the case of Figure 13b, and blue in the case of Figure 13c.
Figure
in in the 15 shows
case ofthe voltage
Figure of Figure
13c. other transmitting
15 shows the coils under of
voltage load (wireless
other power transmission)
transmitting coils under load and
no load (normal)
(wireless condition. and no load (normal) condition.
power transmission)

Figure (a)Initial
13. (a)
Figure 13. Initialposition
positionofofthe
thecoils. XYXY
coils.(b)(b) table
table movement
movement toward
toward nearest
nearest transmitting
transmitting coil.coil.
(c)
(c) Transmitting and receiving coil aligned for wireless power transmission.
Transmitting and receiving coil aligned for wireless power transmission.
Appl. Syst. Innov. 2018, 1, 44 14 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 14 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 14 of 19

Figure
Figure 14.
Figure 14.Transmitting
14. Transmitting coil
Transmitting 11 voltage
coilcoil under
1 voltage
voltage wireless
under
under power
wireless
wireless transmission.
powerpower (For
(For case
transmission.
transmission. scenario
case(For case in
scenario Figure
scenario
in Figure
13).
in Figure 13).
13).

Figure
Figure 15.
Figure 15. Voltage
15. under
under no
Voltage under
Voltage no load
no load (normal)
load (normal) and
(normal) and load
and load (wireless
load (wireless power
(wireless power transmission)
transmission) for
power transmission) for transmitting
for transmitting
transmitting
coil 2.
coil 2.
coil 2.

The
The system
system was was designed
was designed for
designed for the
for the transmission
the transmission power
transmission power of
power of P
of 𝑃 =
𝑃tx == 60
60 W.
60 W.W. Figure
Figure 16 16 shows
shows the the output
output
output
power
power at the load resistance 𝑅
R under
under different
different test
test scenarios.
scenarios.
power at the load resistance 𝑅L under different test scenarios. The misalignment distance between The misalignment distance between
the
the coils
the coils was
coils waschanged
was changedand
changed andthe
and theload
the load power𝑃
loadpower
power 𝑃PLresponse
responsewas
response was
was observed.
observed.
observed. The
The Themaximum
maximum
maximum load
load power
load
powerpower 𝑃
𝑃
was
P measured
was measured to beto 52
be W,
52 which
W, gives
which an
gives efficiency
an efficiency η of η85%.
of
was measured to be 52 W, which gives an efficiency η of 85%. It can be observed in Figure 16 that,
L It
85%.can It be
can observed
be observedin Figure
in 16
Figure that,16
as
as the
that,
theasdistance
the distance
distance of
of misalignment
of misalignment
misalignment between
between the
the transmitting
between the transmitting
transmitting and
and receiving coils
coils increased,
and receiving
receiving coils increased,
increased, the
the time
timethetaken
time
taken
by
taken
by the
thebycharging station
the charging
charging (XY
(XY table)
stationstation (XY to
table) properly
table)
to align
align to
to properly
properly the
the nearest
toalign possible
to the nearest
nearest possible coil
coil also
possible alsocoilincreased. In
also increased.
increased. In the
the
case
case of a misalignment of x = 100 mm, y = 50 mm, it took almost 1.85 s to properly align the coils and
In of
the a
casemisalignment
of a misalignmentof x = 100
of mm,
x = 100y = 50
mm, mm,y = it
50took
mm, almost
it took1.85 s
almost to properly
1.85 s to align
properly the coils
align the
and
transfer
coils andthe
transfer the power
transfer
powerthe with
with designed
power
designedwith efficiency. For
For misalignments
designed efficiency.
efficiency. of
of xx == 75
For misalignments
misalignments 75 mm, of xyy===75
mm, 30
30mm,mm
mm yand and= 30xx ==mm50
50
mm,
and y
x == 10
50 mm, the
y = time
10 mm,to align
the the
time coils
to was
align almost
the coils 1.58
was s and
almost 1.5
mm, y = 10 mm, the time to align the coils was almost 1.58 s and 1.5 s, respectively. In a time ranging s,
1.58 respectively.
s and 1.5 s, In a time
respectively. ranging In a
between
time 1.5
ranging s and
between1.9 s, the
1.5 smisalignment
and 1.9 s, the caused by
misalignment the imperfect
caused
between 1.5 s and 1.9 s, the misalignment caused by the imperfect drone landing was eliminated and bydrone
the landing
imperfect was
drone eliminated
landing and
was
the power
power transmission
eliminated
the and the power
transmission was
was increased
increased to
transmission the
towas maximum
maximumtopower.
theincreased the maximum power.
power.
In
In order
In order to to validate
validate the the hill
hill climbing
climbing algorithm,
algorithm, the
algorithm, the data
data obtained
obtained from
obtained from the the practical
practical test test bench
bench
was used in MATLAB and simulations
simulations were
were carried
carried out
out to
to
was used in MATLAB and simulations were carried out to assure the feasibility of the algorithm. assure
assure the
the feasibility
feasibility of
of the
the algorithm.
algorithm.
Figure
Figure 17
Figure 17 shows
shows the the trajectory
trajectory of
trajectory of the
the XYXY position
position for for one
one transmitting
transmitting coil coil inin aa three-dimensional
three-dimensional
space.
space. It It can
It can be
can be seen
be seen that,
seen that, at
that, at the
at the start,
the start, there
start, there is
there is no
is no voltage
no voltage drop
voltage drop and
drop and the
and the XYXY table
XY table is
table is moved
is moved to
moved to aaa random
to random
random
position.
position. After two random movements, the position with the minimum possible voltage value is
position. After
After two
two random
random movements,
movements, the
the position
position with
with the
the minimum
minimum possible
possible voltage
voltage value
value is
selected and, from
from that position, the hill climbing algorithm is activated
selected and, from that position, the hill climbing algorithm is activated and it keeps finding the best and it keeps finding
finding the best
possible
possible positions
positions to to align
align the
the transmitting
transmitting and and receiving
receiving coilscoils until
until the
the minimum
minimum level level is is achieved.
achieved.
Appl. Syst. Innov. 2018, 1, 44 15 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 15 of 19
Appl. Syst. Innov. 2018, 2, x FOR PEER REVIEW 15 of 19
possible
Once thepositions
minimum to level
alignisthe transmitting
achieved, and receiving
the transmitting coils
and until the
receiving minimum
coils levelproperly
are aligned is achieved.
and
Once the
Once the minimum
minimum level
level is
is achieved,
achieved, the
the transmitting
transmitting and
and receiving
receiving coils
coils are
are aligned
aligned properly
properly and
and
wireless power transmission starts with full efficiency.
wireless power
wireless powertransmission
transmissionstarts
startswith
withfull
fullefficiency.
efficiency.

Figure 16. Load power response for different misalignment cases.

Figure 16. Load power response
Figure 16. response for
for different
different misalignment
misalignment cases.
cases.
The results obtained
The obtained from the the test bench
bench shows that that the proposed
proposed system is is quite efficient
efficient in
The results
results obtained from
from the test
test bench shows shows that the the proposed system
system is quite
quite efficient in in
resolving
resolving the misalignment issue. The power transmission efficiency of 85% is reasonable for
resolvingthethemisalignment
misalignment issue. TheThe
issue. power
powertransmission efficiency
transmission of 85%of
efficiency is reasonable for resonant
85% is reasonable for
resonant inductive-based
inductive-based wireless powerwireless power transmission,
transmission, while also
while also reducing reducing
the need the
for thethe need for the
implementation
resonant inductive-based wireless power transmission, while also reducing need for the
implementation
of complex trackingof complex
and landingtracking and
algorithms landing
for the algorithms
drone. The for the drone. The time to align the
implementation of complex tracking and landing algorithms for time to align
the drone. thetime
The centroid
to alignis also
the
centroid
minimized is also minimized
due minimized
to the use of due to the
thetohill use of the hill climbing algorithm. The drone is free to land
centroid is also due theclimbing
use of the algorithm. Thealgorithm.
hill climbing drone is freeThetodrone
land is anywhere
free to landon
anywhere
charging on charging station and, within a time of just 2 s, the centroid of the coils are aligned
anywherestation and, within
on charging a time
station and,ofwithin
just 2 as, time
the centroid
of just 2ofs,the
thecoils are aligned
centroid of the properly
coils are and the
aligned
properly
power and the power
transmission loss transmission
is mitigated. loss is mitigated.
properly and the power transmission loss is mitigated.

Figure 17. Trajectories

Figure 17. Trajectoriesof
ofhill
hillclimbing
climbingalgorithm.
algorithm.
Figure 17. Trajectories of hill climbing algorithm.
Comparative
ComparativeStudyStudy
Comparative Study
AA comparative
comparative study study ofof the
the proposed
proposed and andpreviously
previously presented
presented solutions
solutions for
for drone
droneimperfect
imperfect
landingA comparative
(misalignment study of is
issues) the proposed
presented in and
this previously
section. The presented
system is solutions for
compared with drone
three imperfect
solutions
landing (misalignment issues) is presented in this section. The system is compared with three
landing
presented (misalignment
in Reference issues) is presented in this section. The system is compared with three
solutions presented in [37–39].
Reference [37–39].
solutions
In presented in Reference [37–39].
In Reference
Reference [37],
[37], the
the authors
authors presented
presented an an approach
approach based
based onon the
the optimal
optimal design
design of of the
the
In Reference
transmitting and [37],
receiving the authors
coils with presented
the goal of an approach
becoming less based on
sensitive tothe
the optimal design
misalignment of
of the
coils.
transmitting and receiving coils with the goal of becoming less sensitive to the misalignment of coils.
transmitting
The and receiving of coils with thestation
goal of becoming less sensitive to the misalignment of coils.
Thesystem
systemwaswascomposed
composed ofaachargingcharging stationwith withmultiple
multiplearrays
arraysofofprimary
primaryor ortransmitting
transmittingcoilscoils
The system
with a was composed
specifically designed of a charging
secondary orstation
receivingwith multiple
coil. The arrays ofcoil
receiving primary
was or transmitting
designed to coils
perfectly
with a specifically designed secondary or receiving coil. The receiving coil was designed to perfectly
with
fit a specifically designed secondary or receiving coil. The receiving coil was designed toschemeperfectly
fitin
inthe
thelanding
landingskidskidof ofthe
thedrone.
drone.The Theauthors
authorsproposed
proposedaatransmitting
transmittingcoilcoiloverlapping
overlapping scheme to to
fit in thecover
entirely landing skid of the drone. The authorsstation.
proposed acalculating
transmitting coil overlapping scheme to
entirely coverthe thecharging
chargingarea areaon onthe
thecharging
charging station.By By calculatingthetheimpedance
impedanceof ofthe
themultiple
multiple
entirely cover
transmitting the charging area on the charging station. By calculating the impedance oftransmission
the multiple
transmittingcoilscoilsand
andchoosing
choosingthe thetransmitting
transmittingcoilcoilwith
with maximum
maximum impedance,
impedance, power
power transmission
transmitting
between coils and choosing the transmitting coil with maximum impedance, power transmission
betweenthe thetransmitting
transmittingand andreceiving
receivingcoils
coilswas
wasachieved.
achieved.
between the transmitting and receiving coils was achieved.
This system works well for a specific type of drone with some specific dimensions; however, if
This system works well for a specific type of drone with some specific dimensions; however, if
the size and dimension of the drone (physical size) changes, a whole new design of the charging
the size and dimension of the drone (physical size) changes, a whole new design of the charging
Appl. Syst. Innov. 2018, 1, 44 16 of 19

This system works well for a specific type of drone with some specific dimensions; however, if the
size and dimension of the drone (physical size) changes, a whole new design of the charging station is
required, and the transmitting coils will need to be readjusted again with some proper overlapping to
cover the charging area properly.
On the other hand, our proposed system has non-overlapped multiple transmitting coils which can
be moved in four directions. Furthermore, the proposed system utilizes the feature of backscattering
which generally happens in wireless communication in order to identify the most closely coupled
receiving coil among them. Even though the closely coupled receiving coil is detected, a further
fine alignment of the centroids of the transmitting coil and the receiving coil is required to obtain
efficient wireless power transmission by moving multiple transmitting coils. The proposed system
is autonomous and totally free from the physical dimensions of the drone; it just needs to detect a
receiving coil. This coil could be placed at any part of the drone where it can come in contact with any
of the multiple transmitting coils placed on the charging station. The drone can land at any part of the
charging station and, within just 1 to 2 s (min to max, depending on the distance of misalignment), the
coils would be properly aligned with a strong coupling factor.
In Reference [33], the impedance of the transmitting coils was monitored, which requires some
current and voltage sensors to provide the measured value at every instance of charging operation.
It increases the complexity of the charging station and requires more electronic components. On the
other hand, in our case, the voltages of the transmitting coils are enough to ensure proper alignment
between coils. Regarding power transfer capabilities and efficiency, in Reference [37], the authors
presented results based on the efficiency of wireless power transmission. The results showed that
the efficiency of the system changed and decreased when there was misalignment between the coils.
The efficiency met the minimum level requirement of 75% in all cases; however, upon misalignment,
there was a reduction in efficiency from 88% (normal; no misalignment) to 83% (100-mm misalignment)
and 78% (200-mm misalignment). In our system, the efficiency will always be the same (normal; no
misalignment) because there is no misalignment and power transmission is done after ensuring this.
In Reference [38], the authors presented a target detection technique based on image processing.
After landing, the center of the coil was aligned with the transmitting coil using some specific color
detection and image-processing scheme. The image-processing algorithm worked by taking the images
using a drone camera and converted RGB color space to HSV color space. After applying some filters,
the red color was detected and considered as a target.
There is always a chance of a failure in such a scheme contingent upon the outer environment and
weather conditions; research work is still being pursued to improve such image-processing techniques.
In our work, the drone is independent of such outer environmental uncertainties and the system works
autonomously with the use of the hill climbing algorithm.
In Reference [39], the authors presented a positioning system using a binary distance laser
sensor and ultrasonic sensors. The system was composed of a charging station comprising a single
transmitting coil and a drone equipped with a single receiving coil. The system worked by detecting the
position of the receiving coil and aligning the transmitting coil with it for wireless power transmission.
The positioning system took almost 5 s to detect and align with the receiving coil. Using sensors,
the system complexity increases; it requires some specific areas for installation of the drone and the
charging station. There is the possibility of errors in installing these sensors properly at the proper
position so that the alignment between coils is always perfect. In our system, there is no physical sensor
that needs to be installed on the drone or charging station. The time it takes to get the best possible
solution is 1 to 2 s (min to max, depending on the distance of misalignment). The algorithm used in
our system is much faster and much more robust, giving accurate results. Moreover, in Reference [35],
there is just a single transmitting coil used to transfer the power. In our case, the charging station is
composed of multiple transmitting coils because using multiple transmitting coils gives the possibility
of designing large charging stations for big drones. If the drone size is big, it will require more space
for landing on the charging station. Using multiple transmitting coils decreases the time required
Appl. Syst. Innov. 2018, 1, 44 17 of 19

to align the coils. Generally, the drone lands at any point on the charging station. In Reference [35],
it was made sure that the drone lands in the range of the sensors to initiate the alignment process;
however, it would be almost impossible to initiate this process of alignment if the drone landed on the
charging station and was out of range of the sensors. For this, in our charging station, multiple arrays
of transmitting coils were installed to allow the drone to land freely anywhere on the charging station
without the possibility of getting out of range.

5. Conclusions
In this work, an efficient wireless power transmission system for drone battery charging was
developed. A charging station with multiple transmission coils was used to transfer power to the
receiving side (drone) to charge the battery. This study aimed to solve the problem caused by uneven
drone landing on a charging station which leads to inefficient wireless power transmission due to
the poor alignment between the transmitting and receiving coils. To solve this problem, a control
mechanism based on the hill climbing algorithm was proposed. The control mechanism was used
to mitigate the uneven landing effect of the drone on a charging station by carefully moving the
charging station to a point where wireless power transmission was maximum. A practical test bench
was developed to test the system feasibility. The power transmission efficiency was 85% and the
test accuracy of the system was 98.8%. It can be observed from the results that the proposed system
performed well compared to previously used techniques, without any need of a physical sensor such
as a position sensor. Also, the use of the hill climbing algorithm gives the system a much faster
response compared to any image-based target detection scheme. It also eliminates the possibility
of getting affected by environmental conditions. The drone simply needs to land on the charging
station, and, within 2 s, wireless power transmission starts at its maximum designed efficiency without
any misalignment.

Future Work
We are working on the real-time implementation of the system. For that, there are several things
which need to be improved and addressed carefully. These include the battery charging issues such as
charging time and the capacity of the battery. Additionally, the wireless power transmission system
will be improved in the future. The efficiency of the system can be improved by looking into some
design parameters. The distance between the transmitting and receiving coils will be improved and
medium-range transmission will be introduced into the system.

Author Contributions: Conceptualization, A.R.; Data curation, M.R.; Formal analysis, M.R.; Investigation, A.R.,
M.T. and S.-H.K.; Project administration, A.R.; Resources, M.T.; Software, A.R., M.T.; Supervision, S.-H.K.;
Validation, A.R. and S.-H.K.; Visualization, S.-H.K.; Writing–review & editing, A.R.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Wang, G.; Liu, W.; Sivaprakasam, M.; Humayun, M.; Weiland, J. Power supply topologies for biphasic
stimulation in inductively powered implants. In Proceedings of the IEEE International Symposium on
Circuits and Systems (ISCAS), Kobe, Japan, 23–26 May 2005; Volume 3, pp. 2743–2746.
2. Hoffmann, G.M.; Rajnarayan, D.G.; Waslander, S.L.; Dostal, D.; Jang, J.S.; Tomlin, C.J. The Stanford Testbed
of Autonomous Rotorcraft for Multi-Agent Control (STARMAC). In Proceedings of the 23rd Digital Avionics
System Conference, Salt Lake City, UT, USA, 28 October 2004.
3. Available online: https://www.icao.int/Meetings/UAS/Documents/Circular%20328_en.pdf (accessed on
22 May 2018).
4. Stafford, J. How a Quadcopter Works | Clay Allen; University of Alaska: Fairbanks, AK, USA, 2014.
Appl. Syst. Innov. 2018, 1, 44 18 of 19

5. Luukkonen, T. Modelling and Control of Quadcopter; Independent Research Project in Applied Mathematics;
Aalto University: Helsinki, Finnland, 2011.
6. Lange, S.; Sunderhauf, N.; Protzel, P. A vision based onboard approach for landing and position control of an
autonomous multirotor UAV in GPS-denied environments. In Proceedings of the International Conference
on Advanced Robotics, Munich, Germany, 22–26 June 2009; pp. 1–6.
7. Bhaskaranand, M.; Gibson, J.D. Low-complexity video encoding for UAV reconnaissance and surveillance.
In Proceedings of the IEEE Military Communications Conference (MILCOM), Baltimore, MD, USA,
7–10 November 2011; pp. 1633–1638.
8. Doherty, P.; Rudol, P. A UAV search and rescue scenario with human body detection and geolocalization.
In Advances in Artificial Intelligence; Springer: Berlin/Heidelberg, Germany, 2007; pp. 1–13.
9. Tomic, T.; Schmid, K.; Lutz, P.; Domel, A.; Kassecker, M.; Mair, E.; Grixa, I.L.; Ruess, F.; Suppa, M.; Burschka, D.
Toward a fully autonomous UAV: Research platform for indoor and outdoor urban search and rescue.
IEEE Robot. Autom. Mag. 2012, 19, 46–56. [CrossRef]
10. DIJ. Available online: https://www.dji.com/ (accessed on 15 June 2018).
11. Mostafa, T.M.; Muharam, A.; Hattori, R. Wireless battery charging system for drones via capacitive power
transfer. In Proceedings of the IEEE PELS Workshop on Emerging Technologies: Wireless Power Transfer,
Chongqing, China, 20–22 May 2017; pp. 1–6.
12. Campi, T.; Cruciani, S.; Feliziani, M.; Maradei, F. High efficiency and lightweight wireless charging
system for drone batteries. In Proceedings of the AEIT International Annual Conference, Cagliari, Italy,
20–22 September 2017; pp. 1–6.
13. Shoki, H. Issues and Initiatives for Practical Deployment of Wireless Power Transfer Technologies in Japan.
Proc. IEEE 2013, 101, 1312–1320. [CrossRef]
14. Finkenzeller, K. RFID Handbook: Fundamentals and Applications in Contactless Smart Cards and Identification;
Wiley: Hoboken, NJ, USA, 2003.
15. Jiang, B.; Smith, J.R.; Philipose, M.; Roy, S.; Rajan, K.S.; Mamishev, A.V. Energy scavenging for inductively
coupled passive rfid systems. IEEE Trans. Instrum. Meas. 2007, 56, 118–125. [CrossRef]
16. Li, P.; Principe, J.; Bashirullah, R. A wireless power interface for rechargeable battery operated neural
recording implants. In Proceedings of the 28th Annual International Conference of the IEEE Engineering in
Medicine and Biology Society (EMBS), New York, NY, USA, 30 August–3 September 2006; pp. 6253–6256.
17. Aldhaher, S.; Yates, D.C.; Mitcheson, P.D. Design and development of a Class EF2 inverter and rectifier for
multi megahertz wireless power transfer systems. IEEE Trans. Power Electron. 2016, 31, 8138–8150. [CrossRef]
18. Sekitani, T.; Takamiya, M.; Noguchi, Y.; Nakano, S.; Kato, Y.; Hizu, K.; Kawaguchi, H.; Sakurai, T.; Someya, T.
A large-area flexible wireless power transmission sheet using printed plastic mems switches and organic
field-effect transistors. In Proceedings of the International Electron Devices Meeting (IEDM), San Francisco,
CA, USA, 11–13 December 2006; pp. 1–4.
19. Aldhaher, S.; Mitcheson, P.D.; Yates, D.C. Load-independent Class EF inverters for inductive wireless
power transfer. In Proceedings of the IEEE Wireless Power Transfer Conference (WPTC), Aveiro, Portugal,
5–6 May 2016; pp. 1–4.
20. Aldhaher, S.; Luk, P.C.-K.; Whidborne, J.F. Tuning Class E inverters applied in inductive links using saturable
reactors. IEEE Trans. Power Electron. 2014, 29, 2969–2978. [CrossRef]
21. Choi, J.; Tsukiyama, D.; Tsuruda, Y.; Rivas, J. 13.56 MHz 1.3 kW resonant converter with GaN FET
for wireless power transfer. In Proceedings of the IEEE Wireless Power Transfer Conference (WPTC),
Boulder, CO, USA, 13–15 May 2015; pp. 1–4.
22. Pinuela, M.; Yates, D.C.; Lucyszyn, S.; Mitcheson, P.D. Maximizing dc-to-load efficiency for inductive power
transfer. IEEE Trans. Power Electron. 2013, 28, 2437–2447. [CrossRef]
23. Low, Z.N.; Casanova, J.J.; Maier, P.H.; Lin, J. Method of load/fault detection for loosely coupled planar
wireless power system with power delivery tracking. IEEE Trans. Ind. Electron. 2010, 57, 1478–1486.
24. Zhong, W.X.; Hui, S.Y.R. Maximum energy efficiency tracking for wireless power transfer systems. IEEE Trans.
Power Electron. 2015, 30, 4025–4034. [CrossRef]
25. Fu, M.; Yin, H.; Zhu, X.; Ma, C. Analysis and tracking of optimal load in wireless power transfer systems.
IEEE Trans. Power Electron. 2015, 30, 3952–3963. [CrossRef]
Appl. Syst. Innov. 2018, 1, 44 19 of 19

26. Yin, J.; Lin, D.; Lee, C.K.; Hui, S.Y.R. A systematic approach for load monitoring and power control in
wireless power transfer systems without any direct output measurement. IEEE Trans. Power Electron. 2015,
30, 1657–1667. [CrossRef]
27. Jang, B.J.; Lee, S.; Yoon, H. HF-band wireless power transfer system: Concept, issues, and design.
Prog. Electromagn. Res. 2012, 124, 211–231. [CrossRef]
28. Si, P.; Hu, A.P.; Hsu, J.W.; Chiang, M.; Wang, Y.; Malpas, S.; Budgett, D. Wireless power supply for implantable
biomedical device based on primary input voltage regulation. In Proceedings of the 2nd IEEE Conference on
Industrial Electronics and Applications, Harbin, China, 23–25 May 2007; pp. 235–239.
29. Low, Z.N.; Chinga, R.A.; Tseng, R.; Lin, J. Design and test of a high-power high-efficiency loosely coupled
planar wireless power transfer system. IEEE Trans. Ind. Electron. 2008, 56, 1801–1812.
30. Available online: http://lasermotive.com/category/uavs/VideoofLaser-PoweredQuadrocopterEnduranceFlight
(accessed on 20 July 2018).
31. Available online: http://www.koreaittimes.com/story/6601/unmanned-aerialvehiclesdeveloped-solely-
koreantechnologies (accessed on 20 July 2018).
32. Uri, K.; Stern, H.; Edan, Y.; Feied, C.; Handler, J.; Smith, M.; Gillam, M. Vision-based autonomous robot
self-docking and recharging. In Proceedings of the IEEE World Automation Congress, Budapest, Hungary,
24–26 July 2006; pp. 1–8.
33. Fetisov, V.; Dmitriyev, O.; Neugodnikova, L.; Bersenyov, S.; Sakayev, I. Continuous monitoring of terrestrial
objects by means of duty group of multicopters. In Proceedings of the XX IMEKO World Congress,
Busan, Korea, 9–10 September 2012.
34. Heeseo, C.; Park, J.; Song, H.; Kim, Y.; Jeong, H. The IoT based automate landing system of a drone for the
round-the-clock surveillance solution. In Proceedings of the IEEE International Conference on Advanced
Intelligent Mechatronics (AIM), Busan, Korea, 7–11 July 2015; pp. 1575–1580.
35. Campi, T.; Cruciani, S.; Feliziani, M. Wireless Power Transfer Technology Applied to an Autonomous Electric
UAV with a Small Secondary Coil. Energies 2018, 11, 352. [CrossRef]
36. Junaid, A.B.; Konoiko, A.; Zweiri, Y.; Sahinkaya, M.N.; Seneviratne, L. Autonomous Wireless Self-Charging
for Multi-Rotor Unmanned Aerial Vehicles. Energies 2017, 10, 803. [CrossRef]
37. Choi, C.H.; Jang, H.J.; Lim, S.G.; Lim, H.C.; Cho, S.H.; Gaponov, I. Automatic wireless drone charging station
creating essential environment for continuous drone operation. In Proceedings of the 2016 International
Conference on Control, Automation and Information Sciences (ICCAIS), Ansan, Korea, 27–29 October 2016;
pp. 132–136.
38. Arakawa, T.; Goguri, S.; Krogmeier, J.V.; Kruger, A.; Love, D.J.; Mudumbai, R.; Swabey, M.A. Optimizing
Wireless Power Transfer from Multi Transmit Coils. IEEE Access 2018. [CrossRef]
39. Casanova, J.J.; Low, Z.N.; Lin, J. Design and optimization of a Class-E amplifier for a loosely coupled planar
wireless power system. IEEE Trans. Circuits Syst. II Express Briefs 2009, 56, 830–834. [CrossRef]
40. Bolaji, A.L.; Femi, B.A.; Shola, P.B. Hill Climbing Algorithm for Solving Patient Admission Scheduling
Problem. Knowl.-Based Syst. 2018. [CrossRef]
41. Merino, L.; Caballero, F.; Martínez-de Dios, J.R.; Ferruz, J.; Ollero, A. A cooperative perception system for
multiple UAVs: Application to automatic detection of forest fires. J. Field Robot. 2006, 23, 165–184. [CrossRef]
42. Choudhary, V.; Singh, S.P.; Kumar, V.; Prashar, D. Wireless Power Transmission: An Innovative Idea. Int. J.
Educ. Plan. Adm. 2011, 1, 203–210.
43. Bao, D.; Gu, J.; Di, Z.; Zhang, T. Optimization of Airport Shuttle Bus Routes Based on Travel Time Reliability.
Math. Prob. Eng. 2018. [CrossRef]

© 2018 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 (http://creativecommons.org/licenses/by/4.0/).