This article has been accepted for publication in a future issue of this journal, but has not been
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics
A Hybrid Transmitter Based Efficiency
Improvement Controller with Full-Bridge Dual
Resonant Tank for Misalignment Condition
Ruikun Mai, Member, IEEE, Zhaotian Yan, Student Member, IEEE,
Yang Chen, Student Member, IEEE, Shunpan Liu, Zhengyou He,Senior Member,IEEE
Abstract—In general, misalignment in a wireless power transfer
system is inevitable, which decreases the efficiency of the system.
In this paper, a full-bridge dual resonant tank (FBDRT) topology
is proposed to improve the system efficiency under the lateral-
misalignment condition of a WPT system. The topology with two
transmitters connected with two separate half bridges is employed
to deliver power to the load. Operating principle, efficiency,
current, power and soft-switching region of the proposed method
are analyzed in detail, and a comparison is drawn in with the
traditional series-series topology. Theoretical analysis shows that
the equivalent impedance of the transmitters can be adjusted,
utilizing the phase-shift modulation. The proposed method holds a
higher efficiency and a wider soft-switching region when the phase
is selected reasonably. A 100W prototype is built to validate the
feasibility of the proposed topology. The efficiency of FBDRT is
always higher than 88.3% with an improvement of 6.8%,
compared with that of the traditional SS topology when the lateral
misalignment increases from 0 to 200mm (0-50% of maximum coil
size).
Index Terms—Wireless power transfer, full-bridge dual
resonant tank, efficiency, soft-switching, phase-shift modulation.
I. INTRODUCTION
W IRELESS power transfer (WPT) based on magnetic
coupling has been successfully employed in many
applications, such as biomedical implants [1], underwater
power supplies [2], mobile phones [3], electric vehicles [4]-[7]
and even trains [8]. Among different performance parameters,
efficiency has drawn the most attention from researchers [9],
[10].
A WPT system, with a variable resistance of the load, a single
transmitter (TX) and receiver (RX) will impact the efficiency
of the system drastically. As a result, many researchers have
focused on altering the equivalent load resistance to the optimal
resistance in real time. For example, a DC-DC converter is used
in the RX [11]-[13], and the equivalent load is changed to
achieve a higher efficiency. However, such methods result in a
complicated system. To avoid an additional DC-DC
This work was supported by the Fundamental Research Funds for the
Central Universities (2682017QY01), National Key Research and
Development Program of China (2017YFB1201002), National Natural Science
Foundation of China (51677155), National Science Fund for Distinguished
Young Scholars (51525702). (Corresponding author: Zhengyou He)
The authors are with the School of Electrical Engineering, Southwest
Jiaotong University, Sichuan 611756, China (e-mail: hezy@swjtu.edu.cn;).
0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
converter, the work [14] proposes a semi-bridge rectifier with
which the maximum efficiency is achieved by pulse width
modulation. Similarly, phase-shift modulation (PSM) of the
rectifier is proposed in [15] to match the optimal load
resistance. Nonetheless, the soft-switching characteristics of
these systems are not studied in detail.
Multiple TXs can simultaneously operate for one RX, and the
efficiency is increased in a specific space [16]-[19]. However,
if the RX is close enough to a specific TX, the two-TX system
might perform with a lower efficiency than the one-TX system
[20]. Current in each TX can be adjusted to increase the
system’s efficiency [21], [22]. Current-controlled transmitting
coils are proposed in [21], where a magnetic field is focused
arbitrarily. Similarly, the current in TXs located at an arbitrary
position can be adjusted by changing the amplitude ratio and
phase difference of two TXs [22]. These results are summarized
in [18], where the two TX system was investigated under
lateral- misalignment condition. The experiment shows that the
coil current ratio should be equal to the coupling coefficient
ratio to obtain the highest efficiency. However, these methods
about multiple TXs systems not only complicate the system but
also increases its cost.
A new dual-active-bridge, series resonant converter
(DBSRC) with a dual tank is proposed in [23], and it beats the
traditional DBSRC by optimizing the tap coefficient x of a
tapped transformer. Such a structure can widen the zero voltage
switching (ZVS) region and improve the efficiency through
PSM, when the tap coefficient x changes. This paper
implements this topology into a WPT system known as a full-
bridge dual resonant tank (FBDRT), and its principle is
transformed into the series-series (SS) topology with one TX
under well-aligned condition. Even if the position of the RX
changes, the proposed system presents a higher efficiency and
a wider soft-switching region than the traditional SS topology.
This paper investigates the theoretical analysis of the optimal
phase shift angle of a two-TX system, where the coupling
between each TX and RX are different. The criterion for
selecting a phase shift angle at a certain position is studied to
maximize efficiency. The topology and the principle of the
proposed FBDRT are discussed in Section II. The system's
modeling and its efficiency are derived and compared to the
traditional SS topology in Section III. Section IV gives the
current and power analysis of the proposed FBDRT. In Section
V and VI, soft-switching behavior and experimental results are
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

given respectively. The conclusion of the proposed work is the coupling model of the FBDRT, where the transmitting coil
drawn in Section VII. Lt is divided into two coils, namely L1 & L2, and the turns of
FBDRT are equal to that of the traditional SS topology, such as
II. TOPOlOGY AND PRINCIPLE ANALYSIS Np1=Np2=NP & Ns=NS. The central tap of the transmitting coil
Fig. 1 (a) shows the traditional SS topology in a WPT system and the midpoint (S) of the two slip capacitors are connected.
which is energized by MOSFETs Q1-Q4 and its antiparallel TX and RX are set to the same resonant frequency f. As
diodes in the TX. On the RX side, there are four diodes D5-D8, shown in Fig. 3, uQ1, uQ2, uQ3 and uQ4 are the driving signals in
which can convert alternating current into direct current, while Q1-Q4. Two MOSFETs in one active bridge have a
CS is a capacitor stabilizing the input voltage. Ct (Cr) and Lt (Lr) complemental signal, with 49.5% duty ratio (because of dead-
are the compensation capacitor and self-inductance of the coil band time). θ is the phase shift angle between two active bridges
in TX and (RX) respectively. The input voltage is Vg, whereas in TX. Therefore, there are ten states in one working period,
the load is RL. which are analyzed as follows:
uQ1
Q1 Q3 D5 D7
Ct iP iS Cr uQ2 t
D
A
CS Lt Lr CS RL t
uQ3
Vg B
C
Q2 Q4 uQ4 θ t
D6 D8

(a) u1 t

ip t
Q1 Q3 D5 D7 u2
C1 ip1
CS1 A is C3 θ
u1 D
L1 t
S L3 u3 CS RL i1,c1 i1,c2
Vg i
ip2 L2
u2 B C
CS2
C2 t
Q2 Q4 D6 D8

(b)
t0 t1 t2 t3 t4 t5 t6 t7 t8 t9 t10
Fig. 3. Operating waveform of the proposed FBDRT.
Fig. 1. (a) Traditional series-series topology; (b) The proposed FBDRT
topology. A. Interval 1: before t0
As shown in Fig. 4 (a), only Q3 is conducted at t0 due to the
Fig. 1 (b) represents the proposed FBDRT topology in a dead-band time. Two resonant currents, ip1 and ip2 are both
WPT, which has two symmetrical TXs with a single converter. negative. The current ip1 begins at point S and goes through C1
The proposed approach requires splitting of the TX into two and D1. The resonant tank 1 regenerates energy to CS1. The
equal parts by connecting a wire in the middle of the TX. u1 (u2) current ip2 starts from point S and goes through CS1, Q3, and C2.
is the voltage between point A (B). The variable ip1, ip2 and is Capacitor CS1 transfers energy to the resonant tank 2. During
are the currents in TX1, TX2, and the RX respectively. The this period, both of ip1 and ip2 will decrease in the negative
positive direction of currents and voltages can be provided in direction.
Fig. 1 (b).
) ) B. Interval 2: t0~t1
Moving (d Moving (d
According to Fig. 4 (b), the direction of the resonant currents
RX NS RX Ns (ip1 and ip2) do not change. The current ip1 decreases to zero and
D
Q1 is switched ON with ZVS. Resonant tank 1 feeds energy to
D
iS C is CS1, and in turn, CS1 delivers energy to tank 2. ip1 and ip2 will
C
B S ip2 B continue to decrease in the negative direction.
iP A TX NP A ip1TX1 Np1 TX 2 Np2
(a) (b) C. Interval 3:t1~t2
As shown in Fig. 4 (c), Q1 and Q3 are conducting during t1
Fig. 2. (a) The structure of coupling model in the traditional SS topology.
(b) The structure of the coupling model in the proposed FBDRT. ~t2. The resonant current ip1 is positive, while ip2 is negative.
The current ip1 begins at point S and goes through CS1, Q1, and
Fig. 2 (a) demonstrates the structure of the traditional C1. Capacitor CS1 transfers energy to the resonant tanks 1 and
coupling model where the number of turns of the transmitting 2. ip1 increases in a positive direction while ip2 continues to
and receiving coils are NP and NS, respectively. Fig. 2 (b) shows decreases in a negative direction.

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

D. Interval 4: t2~t3 ip1 begins from point S and goes through the capacitor CS2, D2,
During period t2 to t3 as shown in Fig. 4 (d), the direction of and C1. Resonant tank 1 regenerates energy to capacitor CS2,
the resonant currents, ip1 and ip2 do not change, ip1 is positive and in turn, CS2 delivers energy to resonant tank 2. As a result,
while ip2 is negative. Q4 is switched ON with ZVS at t2 and the ip1 will decrease while ip2 will increase.
current ip2 decreases to zero at t3. CS1 transfers energy to the G. Interval 6: t5~t6
resonant tank 1 while resonant tank 2 regenerates energy to CS2.
In this interval, resonant current ip1 changes the direction.
ip1 increases in a positive direction while ip2 decreases in a
The current ip2 begins from point S and goes through C1, Q2, and
negative direction.
CS2. Meanwhile, capacitor CS1 delivers energy to resonant tanks
1 and 2.
Q1 Q3 Q1 Q3
C1 C1
H. Interval 6: t6~t7
CS1 A CS1 A
Vg ip1 L1 Vg ip1 L1 Fig. 4 (h) shows the condition from t6 to t7. The direction of
S L3 S L3 resonant currents ip1 and ip2 does not change, and Q3 is switched
ip2 L2 ip2 L2
B B ON with ZVS at t6. Capacitor CS2 transfers energy to the
CS2 CS2
C2 C2 resonant tank 1, and resonant tank 2 regenerates energy to the
Q2 Q4 Q2 Q4
capacitor CS1.
(a) (b)
Similarly, t7~t10 can be analyzed in the same manner as t0 to
t6. be analyzed in the same manner. FBDRT can realize the ZVS
Q1 Q3 Q1 Q3
condition in TX (Q1 to Q4) if the voltages u1 and u2 are ahead of
CS1 A C1 CS1 A C1 the currents, ip1 and ip2.
Vg ip1 L1 Vg ip1 L1
S L3 S L3 III. MODELING AND EFFICIENCY ANALYSIS
ip2 L2 ip2 L2
CS2 B CS2 B A. Modeling of the FBDRT
C2 C2
Q2 Q4 Q2 Q4
The FBDRT’s equivalent circuit is shown in Fig. 5 to analyze
its characteristics. r1, r2, and r3 are the parasitic resistances of
(c) (d)
the TX1, TX2, and RX, respectively. Mij is the mutual
Q1 Q3 Q1 Q3 inductance between Li and Lj (i, j[1,3] and i≠j). For
CS1 A C1 CS1 A C1 simplification purposes, the parasitic resistances of the
Vg ip1 L1 Vg ip1 L1 capacitors and skin effect of the Litz wire are ignored.
S L3 S L3 C1
ip2 L2 ip2 L2
CS2 B CS2 B r1
C2 C2
Q2 Q4 Q2 Q4 M13 r3
U1 ip1 L1
C3
(e) (f) M12
is
L3 RL
Q1 Q3 Q1 Q3
CS1 A C1 CS1 A C1
ip1 ip1 U2 ip2 L2 M23
Vg L1 Vg L1
S L3 S L3
ip2 L2 ip2 L2 r2 C2
CS2 B CS2 B
C2 C2
Q2 Q4 Q2 Q4 Fig. 5. The equivalent circuit of the FBDRT in WPT.

(g) (h) The TX can be seemed as being driven by two AC voltage

Fig. 4 Operating principle of the FBDRT in each interval: (a) Interval 1:
before t0. (b) Interval 2: t0~t1. (c) Interval 3: t1~t2. (d) Interval 4: t2~t3. (e)
Interval 5: t3~t4. (f) Interval 6: t4~t5. (g) Interval 7: t5~t6. (h) Interval 8: t6~t7.
sources U1 and U 2，which have the same amplitude but
different phase shift angles (θ). The following equation shows
the relationship between U1，U 2 and Vg.

E. Interval 5: t3~t4  2Vg

 U = 0 = U 0

1
As shown in Fig. 4 (e), the direction of resonant current ip1
does not change, whereas ip2 becomes positive. The current i2  (1)
U 2 = 2Vg  = U 
begins at point S and goes through C2 and Q4. The capacitor CS1
 
and CS2 transfers energy to the resonant tanks 1 and 2.
The voltages U1，U 2 and the currents I p1，I p2 and Is are
F. Interval 5: t4~t5
assumed to be sinusoidal. According to Kirchhoff’s Voltage
In this interval, both the resonant currents do not change Law, the matrix equation can be obtained as follows:
directions, and Q2 is switched ON with ZVS at t4. The current

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

j M12 − j M13   I p1  U1  d A( f +  ) cos( ) + (2dC − B − Bf ) sin( ) + 2dA

 Z1 = (8)
 j M12 − j M 23    I p 2  = U 2  d [2d cos( ) +  + f ]
2
Z2 (2)
− j M13 − j M 23 Z3   I   0  (8) equals to zero so that η obtains a maximum value (ηmax)
 s
on condition that θ=θmax. θmax is a function of mutual inductance
ω is the angular frequency of the system, and the system’s
and misalignment distance (d), as shown in (9).
frequency is set at f=85kHz. Z1, Z2, and Z3 are the impedances
of each resonant tank while TXs have the same structure in Fig. max = f ( M12 , M13 , M 23 ) = g (d ) (9)
2. The parameters are given in (3). To analyze the characteristics of the system, the parameters
 1 of components are depicted in TABLE I.
 Z1 = Z 2 =r + L +
 jC TABLE I
 (3) PARAMETERS OF FBDRT FOR SIMULATION
 Z =r + R + L + 1 Description Parameter


3 3 L 3
jC3 Input voltage (Vg) /V 60
Resonant frequency(fs) /Hz 85000
The relationship between inductors and capacitors of Loss resistance 1&2(r) /Ω 0.4
resonant tanks are described by: Loss resistance 3(r3) /Ω 0.6
1 Load (RL) /Ω 40
 Li + = 0，i = 1, 2, 3 (4)
jCi
As the RX moves horizontally, the mutual inductances of the
The currents in TXs and RX can be solved according to (2),
FBDRT and the traditional SS topology will change, which can
as follows:
be obtained by ANSYS MAXWELL. Results show that the
 U variation of M13 and M12 are very small compared to that of M23
I p1 =   [(c + dj ) sin( ) + ( d − cj ) cos( ) + ( +  j )] so that M13 and M12 can be viewed as constants to simplify the
 U operation. M can be regarded as the sum of M13 and M23. , Fig.
I p 2 =  [(e + fj ) sin( ) + ( f − ej ) cos( ) + ( d − cj )] (5) 6 shows the variations of the mutual inductance and the
  coupling coefficient, when the RX moves horizontally towards
I =  [( a + bj ) sin( ) + (b − aj ) cos( ) + ( g + hj )]
U
TX1, kij = M ij / Li Lj , i, j=1, 2, 3).
 s 
where the symbols are given in the appendix. 45 0.45
M12 k12
35 M13 0.35 k13
M23 k23
B. System Efficiency Analysis M k
M/µH

25 0.25
The power equations of the FBDTR can be written as (6), in

k
15 0.15
which Pin1 and Pin2 (Pout) are the power delivered (received) by
TX1 and TX2 (RX), respectively. 5 0.05

 1 * U
2 -5 0 40 80 120 160 200
-0.050 40 80 120 160 200
=
 in1 2
P  Re(U  I ) = [c sin( ) + d cos( )+ ] d/mm d/mm
2
1 p1

 (a) (b)
 1 * U
2

P =
 in 2 2 Re(U  I ) = [ d cos( ) − c sin( )+f ] (6) Fig. 6. (a) The mutual inductance of the FBDRT (M12, M13, and M23) and the

2 p2 traditional SS topology (M). (b) The coupling coefficient of the FBDRT (k12,
 2
k13, and k23) and the traditional SS topology (k).
P = 1 I  I *  R = U [ A sin( ) + B cos( ) + C ]
2

 out 2 s s L 2 System’s efficiency versus angle θ with different d is shown

in Fig. 7. FBDRT can maintain maximum efficiency with
Where,
different misalignment distances by using the optimized θ. At a
 A = 2rRL 3 M 12 ( M 132 − M 232 )

B = 2 RL M 13 M 23 (r +  M 12 )
2 2 2 2
100 d=0mm
C = RL 2 ( M 132 + M 232 )(r 2 +  2 M 12 2 ) d=60mm
 d=120mm
d=200mm
80
The input power can be written as Pin=Pin1+Pin2. Therefore,
Efficiency/%

the system efficiency (η) can be obtained as

Pout A sin( ) + B cos( ) + C
= = (7)
60
Pin 2d cos( )+ +f
According to (7), η has no relationship to the input voltage θmax2 θmax4
40
Vg while it is a function against the mutual inductances M12, θmax1 θmax3
M13, M23, and θ. There exists a maximum value of η at a certain
misalignment while θ is varying from 0 to 2π. (8) takes the -200 -100 0 100 200
θ/deg
derivative of η (θ).
Fig. 7. Efficiency varying with θ at the different level position (RL=40).

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

certain misalignment, the efficiency of the system is related to shown in Fig. 10. The mutual inductance will fluctuate in the
θ and load RL. Fig. 8. shows the efficiency variation when the process of moving so that the PSM should be adopted to
misalignment distance is 60mm (k13/k23=1.8), and there exist an distribute the energy transferred by two TXs which undertake
optimal load RL and angle θ. the same energy transmission in condition 1. The TX1 will
M13 will be equal to M23, on condition that d=0, so that the work as the main energy transmission channel with RX close to
current in TXs will be same. Therefore, the middle wire that the TX1 and vice versa. However, such method has a tolerance
connects the point S with the transmitting coil can be taken limitation, as shown in TABLE II.
away to make it like a traditional SS topology. The parameters RX
of the main components are shown in (10). d=0
d<a
TX 1 TX 2
90 a a Process A a a
Condition 1 Condition 2
80
Process B
70
100
60 d >a d=a
80
Efficiency/%

60
50 Process C a a
a a

40 40
Condition 4 Condition 3

20 30 Fig. 10 The condition and the progress of moving the RX to the TX1.
70 50 300
80 60 20
0 90
200
0 TABLE II
25 100 10
50 75
TOLERANCE LIMITATION OF FBDRT
100 0 RL/Ohm Mutual inductance
θ/deg Process State
M12 M13 M23
Fig. 8. Efficiency varying with θ and the load RL when d=60mm (k13/k23=1.8). A Constant Constant Decrease Valid
B Constant Constant Decrease Valid
C Constant Decrease Decrease Invalid
M = M 13 + M 23

M 12 = 0 (10) In order to verify the range of misalignment, the mutual
 AB = 0 inductance at a different position is taken into (7), and the
efficiency curves can be drawn as Fig. 11, where the loads are
Taking these parameters into (7), the paper gives the 40 Ω, 80 Ω, and 120 Ω respectively.
efficiency of the traditional SS topology,
 M RL
2 2 100
Pout
t = = (11)
( r3 + RL )[2( r3 + RL ) r +  M ] 90
2 2
Pin
80
Fig. 9 shows the maximum efficiency curve of FBDRT and
traditional SS topology with different misalignment conditions, 70
Efficience/%

where the loads are 40 Ω, 80 Ω, and 120 Ω, respectively. 60

FBDRT can maintain a higher efficiency utilizing PSM, while
50 FBDRT RL =40
the efficiency of traditional SS topology decreases rapidly. FBDRT RL =80
40 FBDRT RL =120
S-S RL =40
30 S-S RL =80
90 S-S RL =120
20
0 120 40 160 200
80 240
d/mm
80
Fig. 11 Efficiencies of FBDRT under misalignment condition (a=200mm).
Efficiency/%

70 According to Fig.11, the range of misalignment can be

FBDRT RL =40
FBDRT RL =80 summarized as −a  d  a , where a is the size of the
FBDRT RL =120
60
S-S RL =40 transmitting coil (a=200mm).
S-S RL =80
S-S RL =120 D. The effect of M12 on efficiency
50
0 40 80 120 160 200 The mutual inductance between TXs (M12) will influence the
d/mm
efficiency. In order to illustrate the relationship between the
Fig. 9. Efficiencies of FBDRT compared with traditional SS topology.
system’s efficiency and M12, the paper introduces the BPP
C. The range of misalignment structure which consists of two identical, partially overlapped
It is necessary to illustrate the range of misalignment and mutually decoupled coils [24]. The mutual decoupling of
tolerance, and the event can be analyzed as three processes, as the coils constituting the BPP is achieved by adjusting the

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

overlap of the two coils, and the mutual inductance between the with different misalignment conditions where M12 is 0 µH, 2.5
two coils depends on the extent of overlap, as shown in Fig. 12. µH, and 5 µH, respectively. Fig. 15 (b) shows the efficiencies
l is the length of the overlap of the two coils. There is no of FBDRT varying with M12 at three positions, and efficiency
change in coil size so that the fluctuation of M13 and M23 is very is improved significantly when M12 increases.
small when the transmitting coils are placed in Fig. 12. 100 100

80 80

Efficiency/%
M12=5µH M12=2.5µH M12 =0µH

Efficiency/%
60 60
200mm

40 40
d=0mm
M12=5µH d=80mm
20 M12=2.5µH 20
d=160mm
200mm 200mm l=42mm l=65mm M12=0µH
0 00
(a) (b) (c) 0 40 80 120 160 200 1 2 3 4 5
d/mm M12/uH
TX1 TX2 (a) (b)
Fig. 12. The value of M12 when two transmitting coils are placed in different
positions. Fig. 15 (a) Efficiencies of FBDRT under the different value of M12 (RL=40Ω).
(b) Efficiencies of FBDRT varying with M12 at three positions (RL=40Ω).
The previous theoretical derivation was based on no
overlapping between the transmitting coils, as shown in Fig. 12
(a). Take the value of M12 into equation (7), and the efficiency IV. CURRENT AND POWER
versus phase shift angle θ at different misalignment conditions A. Current in FBDRT and traditional SS topology
can be shown as follows. According to (5), the RMS value of currents in FBDRT can
be derived as follows.

100 100
M12=5µH M12=5µH I p1
I p1 = 2
M12=2.5µH 90 M12=2.5µH
90
M12=0µH 80 M12=0µH
Efficiency/%

Efficiency/%


80 70
70 60 Ip2
60
50 I p2 = 2 (12)

40
50 30
40
20 I s = Is
30
10
0
 2
-150 -100 -50 0 50 100 150 -150 -100 -50 0 50 100 150
θ/deg θ/deg Similarly, according to equation (10) and Kirchhoff’s
(a) (b)
Voltage Law, the RMS currents in traditional SS topology can
Fig. 13. (a) Efficiency varying with θ and M12 (d=0mm). (b) Efficiency
varying with θ and M12 (d=100mm). be described as,
 2U ( r3 + RL )
According to Fig. 13 (a), the maximum efficiency is the same  I P = 2( r + R ) r +  2 M 2
under well-aligned condition (d=0mm) while it will drop down

3 L
(13)
2 MU
with M12 decrease when the lateral misalignment occurs, as  IS =
shown in Fig. 12 (b). Namely, the system will get a better  2( r3 + RL ) r +  M
2 2

condition if M12 is greater.

Furthermore, the extreme value of efficiency is unrelated to IP and IS are the RMS current of the transmitting coil and the
the phase shift angle θ, when two transmitting coils are receiving coil in the traditional SS topology. In both traditional
decoupled (M12=0). Therefore, the efficiency will get its SS topology and proposed FBDRT topology, the currents in
maximum value when θ=0, as shown in Fig. 14. transmitters and receiver will increase under the misalignment
condition because of the fluctuation of mutual inductance, as
100 shown in Fig. 16.
d=0mm
90 d=80mm 15 4
IP
80 d=160mm Ip1 IS
3.5 Is
Efficiency/%

70 Ip2
10
Current/A

Current/A

60 3
50
40 2.5
5
30 2
20
10 00 40 80 120 160 200 1.5 0 40 80 120 160 200
0 d/mm d/mm
-200 -150 -100 -50 0 50 100 150 200 (a) (b)
θ/deg Fig. 16 (a) The current variation of transmitters in the proposed FBDRT and
Fig. 14 Efficiency varying with θ when transmitting coils are decoupled the traditional SS topology. (b) The current variation of receivers in the
(M12=0) when the lateral misalignment occurs. proposed FBDRT and the traditional SS topology.

Fig. 15 (a) shows the maximum efficiency curves of FBDRT Fig. 16 (a) shows the currents variation of transmitters in

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

FBDRT and traditional SS topology, respectively, while Fig. 16 ZVS condition of leg A and B.
(b) shows that of receivers. When the RX is close to the TX1, With the increase of d, ZVS condition changes due to the
the TX1 will serve as the main energy transmission channel so variation of the mutual inductance and θ. The mutual
that the current in the TX1 will increase through the proposed inductance (Fig.6) can be substituted into (12), and the ZVS
PSM. However, the currents in the traditional SS topology are condition is recalculated. Fig. 19 (a) shows the relationship
uncontrollable, and the currents variation of the traditional SS between the ZVS region of Leg A and the maximum efficiency
topology are bigger than that of the proposed FBDRT topology curve at different positions. Black lines are the boundaries of
apparently. Namely, the proposed FBDRT has better constant ZVS region, while the red line is the maximum efficiency curve
current characteristics, and the current stress can be lightened. at a different position.
200 200
B. Power in FBDRT
160 160
According to equation (6), Pin1, Pin2 and Pout of FBDRT can
120 120

d/mm
be shown in Fig. 17. It is clear that the variation trend of power

d/mm
A B
is the same as that of currents. 80 80
40 40
600
Pin1 0 0
500 Pin2 -100 -50 0 50 100 -100 -50 0 50 100
Pout θ/deg θ/deg
Power/W

400 ZVS region Efficiency ZVS region Efficiency

300 (a) (b)

200 Fig. 19. ZVS condition of the proposed FBDRT: (a) ZVS for leg A when level
position varies from 0 to 200mm (the region A is the ZVS region for leg A); (b)
100 ZVS for leg B (the region B is the ZVS region for leg B)
00 40 80 120 160 200
d/mm Similarly, region B is also a ZVS region and black lines are
Fig. 17 The variation of the power in FBDRT. the boundaries in Fig. 19 (b). It can be found that both legs A
and B have different ZVS regions. The conclusion of the ZVS
V. ANALYSIS OF SOFT-SWITCHING of FBDRT is drawn in TABLE III.
TABLE III
Both the proposed FBDRT topology and the traditional SS ZVS CONDITION OF FBDRT IN DIFFERENT POSITION
topology have two bridge legs in the inverter. Each bridge leg 0-160 160-200
d/mm
in TX has an equivalent circuit. Fig. 18 (a) gives the ZVS (0-40%) (40%-50%)
condition for one leg, whereas Vn (t) is the voltage of the point ZVS condition Leg A &Leg B Leg B
N. In [25], the ZVS condition for one leg is as iX1>0, iX2<0.
u1
i1,c1 VI. EXPERIMENT RESULTS
As shown in Fig. 20, a prototype WPT system is set up to
QH Vn(t)
iX1 t verify the proposed topology according to the parameters in
in
N u2 TABLE . SiC MOSFETs (C2M0160120D) are chosen as the
t i1,c2
iX2 switches in the inverter. There is only one inverter with a single
QL in(t)
Vn DC input source (IT6526D) in TX. The gate drives the signal,
t

and the PSM is generated by the STM32F103. The RX coil is
connected to the rectifier and an electronic load (IT8818BS)
(a) (b) serves as a load resistor. The air gap between the TX and RX is
Fig. 18. (a) Voltage and current waveform of one leg. (b) Voltage and current 6.0cm, while RX is covered with ferrite cores.
waveform of TX1 and TX2. TABLE IV
PARAMETERS OF FBDRT FOR EXPERIMENT
The proposed FBDRT has two TXs. The relationship Description Parameter
between the voltage and current in two TXs is shown in Fig. 18 Input voltage (Vg) / V 60
Resonant frequency (fs) / Hz 85000
(b). It can be found that if ui is ahead of ipi (i=1, 2), leg A and B Capacitor 1 (C1) / nF 60.8
can work in a ZVS condition. Capacitor 2 (C2) / nF 63.3
Capacitor 3 (C3) / nF 20.2
 U1
 1 = arg( I )  0
Capacitor S1 (CS1) / µF 470
Capacitor S2 (CS2) / µF 470
Load (RL) / Ω

p1 40
(14)
U2 Inductance 1 (L1) / µH 57.7
 2 = arg( )  0 Inductance 2 (L2) / µH 55.4
 Ip2 Inductance 3 (L3) / µH 173.5

The phase angles φ1 and φ2 between the voltage and the

current in two TXs can be expressed as (12), which gives the The appearance of TX and RX is shown in Fig. 21 (a). The
dimensions of TXs and RX is 20cm×20cm and 20cm×40 cm

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

respectively. The turns of TX and RX coils are equal, i.e. 100

Np1=Np2=Ns=17. The measured parasitic resistance is 0.4 Ω of 90
one transmitting coil and 0.6Ω of the receiving coil. The coil is 80
wounded by Litz wire of 3 mm×600 strands. The mutual

Efficience/%
70
inductance with misalignment of FBDRT is shown in Fig. 21
60
(b).
50
RX Efficiency with PSM
Rectifier C3 40 Efficiency without PSM
C1 30 Efficiency in traditional S-S
TX2
20
Inverter 0 40 80 120 160 200
CS1 TX1 C2 d/mm
(b)
CS2
Fig. 22 (a) The Efficiency Curves of FBDRT with RL=40Ω. (b) The efficiency
STM32 variation about FBDRT with PSM, FBDRT without PSM and traditional SS
topology at different positions.

to decouple the two coils completely in the experiment. The

Fig. 20 The experimental setup efficiency versus phase shift angle θ is measured where M12 is
0.22 µH, 2.45 µH, and 4.82 µH respectively, as shown in Fig.
RX N3
23.
C 20cm 25
Mutual inductance/µH

D 20
M13 100 95
M12=4.82 M12=0.22
40cm 15 M23 95 M12=2.45
10 M12 90 90
Efficiency/%

Efficiency/%
TX1 N1 20cm 20cm 85
5 80 85
TX 2 N2 0 75
M M12=4.82
70 M12=2.45 80
A B -5
0 40 80 120 160 200 65 M12=0.22
d/mm 60 0 10 75 0 10 20 30 40 50 60 70 80 90
20 30 40 50 60 70 80 90
(a) (b)  /deg  /deg
(a) (b)
Fig. 21 (a) Coils in the TXs and the RX of FBDRT (b) The FBDRT’s mutual
inductance with misalignment. Fig. 23 (a) Efficiency varying with θ and M12 (d=0mm). (b) Efficiency
varying with θ and M12 (d=100mm).

Fig. 22 (a) gives the efficiency curves of FBDRT with RL=40Ω.

Fig. 24 shows the maximum efficiency curves of FBDRT
The RX is placed at 0mm, 60mm, 120mm and 200mm from
with different misalignment conditions where M12 is 0.22 µH,
TXs, while θ changes from 0 to π/2. The maximum efficiency
2.45 µH, and 4.82 µH respectively. The efficiency will drop
is different in every curve. Fig. 22 (b) gives the maximum
sharply if M12 decreases.
efficiency curve of FBDRT with PSM, without PSM in FBDRT
100
and with traditional SS topology. The performance of FBDRT M12=4.82
with PSM is better than that of the traditional SS topology. For 95 M12=2.45
M12=0.22
the FBDRT, the maximum efficiency is 94.13% when d=0 and
90
Efficience/%

is higher than 88.3%, provided the RX moves from 0 to 200mm.

The BPP structure is introduced in this experiment, and the 85
mutual inductance between the two coils (M12) depends on the
extent of overlap, as shown in Fig. 12. However, it is impossible 80
100
75
90
70
80 0 40 80 120 160 200
d/mm
Efficience/%

70 d=0mm Fig. 24 Efficiencies of FBDRT under the different value of M12 (RL=40Ω).
60 d=60mm
d=120mm
50 d=200mm Fig. 25 shows the power variation of TX1, TX2, input, and
40 output in FBDRT. When the RX is close to the TX1, the TX1
30
will serve as the main energy transmission channel so that the
current in the TX1 will increase through the proposed PSM.
20
0 10 20 30 40 50 60 70 80 90 The experimental waveforms of voltages u1, u2, and currents
θ/deg
ip1, ip2 can be seen in Fig. 26, with the operating conditions of
(a)
Vg=80V, RL=40Ω, and d=120mm. Currents in TX1 and TX2 are

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

unbalanced when θ=0°and the switching loss is high. Leg A

ZVS
will realize ZVS condition and the switching loss of leg B will u1 ip1 Leg A
be smaller when θ=55°. The efficiency can reach its maximum
value of 90.2% at θ=80°. Leg A and B can both realize ZVS at
maximum efficiency.
400 Efficiency=90.2%
Pin2 θ=80°
350 Leg B
Pin1 ip2
Pin u2 ZVS
300 Pout
250
Power/W

200
150
100 (c)
50 Fig. 26 (a) The experimental waveforms of voltages u1, u2, and currents ip1, ip2
when d=120mm and θ=0°. (b) The experimental waveforms of voltages u1, u2,
0 and currents ip1, ip2 when d =120mm and θ=55°. (c) The experimental
0 20 40 60 80 100 120
Fig. 25 The variation of the power in FBDRT. waveforms of voltages u1, u2, and currents ip1, ip2 when d=120mm and θ=80°.

The system is inevitably disturbed in practical applications, Though the RMS values of U1 and U2 are not always the
which causes the voltage fluctuation at the point S, and the RMS same, the system will not crash if the S-point voltage fluctuates.
values of U1 and U2 are not the same. Therefore, it is necessary It is necessary to balance the voltage on the two capacitors (CS1
to analyze the voltage fluctuation at the point S. & CS2) so that the proposed theoretical derivation can work.
The initial voltages of the CS1 and CS2 are set to the different
values, and the voltage of CS2 is equal to that of point S. Fig. 27 UCS2 (US) d=100mm
shows the voltages fluctuation of UCS1, UCS2, and US. The initial UCS1
voltages of CS1 (CS2) are set to 30V (30V) and 20V (40V) B
respectively (between point O and point A). The system starts
running at point A, and the voltage at point S (US) does not O A
change within the time of observation (128s & 132s).
132s
ip1 u1 Leg A
UCS2 (US)=30V
UCS1=30V
(a)
Efficiency=85.4% θ=0° Leg B
u2 No ZVS
ip2 UCS2 (US) d=100mm
UCS1

UCS2 (US)=40V
O
(a) B
A
UCS1=20V 128s
ip1 Leg A
u1

Efficiency=88.8% θ=55° (b)

No ZVS
ip2 Leg B Fig. 27 (a) The voltages fluctuation value of UCS1, UCS2, and US
u2
(UCS1=UCS2=30V). (b) The fluctuation value of the UCS1, UCS2, and US
(UCS1=20V, UCS2=40V).

The method which can balance the voltages at point S can be

shown as Fig. 28. The two resistors are paralleled at CS1 and
(b) CS2, and the resistance of RS1 and RS2 are equal. The paralleling
resistors provide an energy path when the voltage fluctuates

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

between the two capacitors (CS1 and CS2). always higher than 88.3% with an improvement of 6.8% when
the RX moves from 0 to 200mm (0-50% of maximum coil size).
S1 Q1 Q3 APPENDIX
RS1 CS1 C1
A Symbols in Section III can be written as follows.
S U1 L1  = [(r + R )( 2 M 2 + r 2 ) + r 2 ( M 2 + M 2 )]2 + 4 6 M 2 M 2 M 2
 3 L 12 13 23 12 13 23
Vg 
C2 L2  = [ 2
M 23
2
+ r ( r3 + RL )][ r 2
( M 13
2
+ M 23
2
) + ( r 2
+  2
M 12
2
)( r3 + RL )]
CS2 U2 
  = 2 M12 M13M 23[ M 23 + r (r3 + RL )]
RS2 B 3 2 2

Q2 Q4 
S2
(A.1)

a = −r M 23[r ( M13 + M 23 ) + (r + M12 )(r3 + RL )] − 2 M12 M13 M 23
Fig. 28 The improved topology to balance the voltage of CS1 and CS2. 2 2 2 2 2 2 5 2 2

b =  M M [r ( M
13 − M 23 ) + ( r +  M12 )( r3 + RL )]
2 2 2 2 2 2 2
 12 13
The waveform can be shown as Fig. 29. During point O and 
c =  M (r + R )[r 2 ( M 2 + M 2 ) + (r 2 +  2 M 2 )(r + R )] + 2 5 M M 2 M 2
point A, the initial voltage of CS1 and CS2 are 40V and 20V  12 3 L 13 23 12 3 L 12 13 23

d = − M13 M 23[r ( M13 + M 23 ) + (r −  M12 )(r3 + RL )]
2 2 2 2 2 2 2
respectively. At point A, the system begins to operate, and the
voltage of the two capacitors are unequal. At point B, the (A.2)
switches S1 and S2 turn on so that the voltage at point S can be e = −2 3M12 M13M 23[ 2 M132 + r (r3 + RL )]

balanced. In the experiment, RS1 and RS2 are all 5.1kΩ, and CS1  f = [ 2 M132 + r (r3 + RL )][r 2 ( M132 + M 232 ) + (r 2 +  2 M122 )(r3 + RL )]
and CS2 are 470 µF. The power loss of the resistance is about 
 g =  M12 M 23[r ( M 23 − M13 ) + (r +  M12 )(r3 + RL )]
2 2 2 2 2 2 2
0.35W in the experiment, which shows the little effect on the 
h = r M13[r ( M13 + M 23 ) + (r + M12 )(r3 + RL )] + 2 M12 M13M 23
2 2 2 2 2 2 5 2 2
system’s efficiency (the output power of the system is above
100W). (A.3)

d=100mm
O A B C
102s
37s 40s
UCS2 (US)
UCS1
UCS2 (US)=20V
UCS1=40V
Fig. 29 The voltage fluctuation of CS1 and CS2 when d=100mm.
The voltage balance time is related to the value of CS1 (CS2)
and RS1 (RS2), and the time constant  can be defined as follows,
 =RS1  CS1 =RS2  CS2 (15)
The voltage balance time will be longer if  gets a large value.
In this experiment, the voltage balance time is about 5 second.
VII. CONCLUSION
In this paper, FBDRT topology is proposed to improve
efficiency when lateral-misalignment occurs in a WPT system.
According to the theoretical analysis, PSM is applied to TXs,
and ZVS condition can be achieved. The equivalent impedance
of two TXs is adjusted at different positions, and there is an
obvious improvement in efficiency compared to the traditional
SS topology. The proposed method is validated by a 100W
experimental prototype. The experimental results show that the
efficiency in comparison to the traditional SS topology is
REFERENCES
[1] D. Ahn and S. Hong, “Wireless power transmission with self-regulated
output voltage for biomedical implant,” IEEE Trans. Ind. Electron., vol.
61, no. 5, pp. 2225-2235, May. 2014.
[2] Z. Cheng, Y. Lei, K. Song et al., “Design and loss analysis of loosely
coupled transformer for an underwater high-power inductive power
transfer system,” IEEE Transactions on Magnetics, vol. 51, no. 7, pp. 1-
10, Jul. 2015.
[3] Y. Jang and M. M. Jovanovic, “A contactless electrical energy
transmission system for portable-telephone battery chargers,” IEEE Trans.
Ind. Electron., vol. 50, no. 3, pp. 520-527, Jun. 2003.
[4] C. Zheng, J. Lai, R. Chen et al., “High-efficiency contactless power
transfer system for electric vehicle battery charging application,” IEEE
Journal of Emerging and Selected Topics in Power Electronics, vol. 3, no.
1, pp. 65-74, Mar. 2015.
[5] S. Y. Choi, B. W. Gu, S. Y. Jeong et al., “Advances in wireless power
transfer systems for roadway-powered electric vehicles,” IEEE Journal of
Emerging and Selected Topics in Power Electronics, vol. 3, no. 1, pp. 18-
36, Mar. 2015.
[6] J. T. Boys and G. A. Covic, “The inductive power transfer story at the
University of Auckland,” IEEE Circuits and Systems Magazine, vol. 15,
no. 2, pp. 6-27, May. 2015.
[7] J. M. Miller, P. T. Jones, J. Li et al., “ORNL experience and challenges
facing dynamic wireless power charging of EV's,” IEEE Circuits and
Systems Magazine, vol. 15, no. 2, pp. 40-53, May. 2015.
[8] J. H. Kim, B. Lee, J. Lee et al., “Development of 1-MW inductive power
transfer system for a high-speed train,” IEEE Trans. Ind. Electron., vol.
62, no. 10, pp. 6242-6250, Oct. 2015.
[9] A. Berger, M. Agostinelli, S. Vesti et al., “A wireless charging system
applying phase-shift and amplitude control to maximize efficiency and
extractable power,” IEEE Trans. Power Electron., vol. 30, no. 11, pp.
6338-6348, Nov. 2015.
[10] Zhao-Hong Ye, Yue Sun et al., “Energy efficiency analysis of u-coil
wireless power transfer system,” IEEE Trans. Power Electron., vol. 31,
no. 7, pp. 4809-4817, Sep. 2015.
[11] H. Li, J. Li, K. Wang, W. Chen, and X. Yang, “A maximum efficiency
point tracking control scheme for wireless power transfer systems using
magnetic resonant coupling,” IEEE Trans. Power Electron., vol. 30, no.
7, pp. 3998-4008, Jul. 2015.
[12] D. Kobayashi, T. Imura, Y. Hori, “Real-time coupling coefficient
estimation and maximum efficiency control on dynamic wireless power
transfer for electric vehicles,” in Proc. 2015 IEEE PELS Workshop Emerg.
Technol, Wireless Power, Jun. 2015, pp. 1-6.

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2019.2916231, IEEE
Transactions on Power Electronics

[13] W. X. Zhong and S. Y. R. Hui, “Maximum energy efficiency tracking for Yang Chen (S’17) received the B.Sc. degree in
wireless power transfer systems,” IEEE Trans. Power Electron., vol. 30, electrical engineering and automation from the
no. 7, pp. 4025-4034, Jul. 2015. School of Electrical Engineering, Southwest
[14] T. Diekhans and R. W. De Doncker, “A dual-side controlled inductive Jiaotong University, Chengdu, China, in 2015,
power transfer system optimized for large coupling factor variations and where he became a mater student since September
partial load,” IEEE Trans. Power Electron., vol. 30, no. 11, pp. 6320-6328, 2015. He was selected to continue his study as s
Nov. 2015. doctoral candidate in September 2016. In 2018, he
[15] A. Berger, M. Agostinelli, S. Vesti et al., “A wireless charging system received the Scholarship under the State Scholarship
applying phase-shift and amplitude control to maximize efficiency and Fund of China Scholarship Council, and became a
extractable power,” IEEE Trans. Power Electron., vol. 30, no. 11, pp. joint Ph.D. student with the Future Energy
6338-6348, Nov. 2015. Electronics Center (FEEC), Virginia Tech, Blacksburg, VA, USA.
[16] R. Johari, J. Krogmeier and D. Love, “Analysis and practical His research interest focuses on wireless power transfer, especially on
considerations in implementing multiple transmitters for wireless power compensation topology and misalignment tolerance improvement.
transfer via coupled magnetic resonance,” IEEE Trans. Ind. Electron., vol. Shunpan Liu received the B.S. degree in electrical
61, no. 4, pp. 1774-1783, Apr. 2014. engineering and automation from the School of
[17] F. Jolani, Y.-Q. Yu and Z. Chen, “A planar magnetically-coupled resonant Electrical Engineering, Southwest Jiaotong
wireless power transfer using array of resonators for efficiency University, Chengdu, China, in 2018, where he is
enhancement,” in Proc. IEEE Int. Microw. Symp., 2015, pp. 1-4. currently working toward the Ph.D. degree.
[18] S. Huh, D.Ahn, “Two-transmitter wireless power transfer with optimal His main research interest focuses on wireless
activation and current selection of transmitters,” IEEE Trans. Power power transfer.
Electron., vol. 33, no. 6, pp. 4957-4967, Jun. 2018.
[19] D. Ahn, S. Hong, “Effect of coupling between multiple transmitters or
multiple receivers on wireless power transfer,” IEEE Trans. Ind. Electron.,
vol. 60, no. 7, pp. 2602-2613,Jul. 2013.
[20] I.-J. Yoon and H. Ling, “Investigation of near-field wireless power
transfer under multiple transmitters,” IEEE Antennas Wireless Propag. Zhengyou He (M’10-SM’13) received the B.S. and
Lett.,vol. 10, pp. 662-665, Jun. 2011. M.S. degrees from Chongqing University,
[21] B. Choi, J. Kim, J. Cheon, and C. Rim, “Synthesized magnetic field Chongqing, China, in 1992 and 1995, respectively,
focusing using a current-controlled coil array,” IEEE Magn. Lett., vol. 7, and the Ph.D. degree from Southwest Jiaotong
Feb. 2016, Art. no. 6501504. University, Chengdu, in 2001. Since 2002, he has
[22] P. Kong and H. Ku. “Efficiency optimizing scheme for wireless power been a Professor with the School of Electrical
transfer system with two transmitters,” Electron. Lett., vol. 52, no. 4, pp. Engineering, Southwest Jiaotong University. He was
310-312, Feb. 2016. a Visiting Scholar with Cornell University, NY,
[23] J. Wu, Y. Li et al., “A new dual-bridge series resonant DC-DC converter USA, from 2008 to 2009. His research interests are
with dual tank,” IEEE Trans. Power Electron., vol. 33, no. 5, pp. 3884- in the area of signal processing and information
3897, May 2018. theory and its application in electrical power systems,
[24] A. Zaheer, Grant A. Covic, and D. Kacprzak, “A bipolar pad in a 10-kHz and the application of wavelet transforms in power systems.
300-W distributed IPT system for AGV applications,” IEEE Trans. Ind.
Electron., vol. 61, no. 7, pp. 3288-3301, July 2014.

Ruikun Mai (M’14) received the B.Sc. and Ph.D.

degrees in electrical engineering from the School of
Electrical Engineering, Southwest Jiaotong
University, Chengdu, China, in 2004 and 2010,
respectively. He is currently a Professor with the
School of Electrical Engineering, Southwest Jiaotong
University, Chengdu, China. His research interests
include wireless power transfer and its application in
railway systems, power system stability and control.

Zhaotian Yan received the B.Sc. degree in electrical

engineering and automation from the School of
Electrical Engineering, Southwest Jiaotong
University, Chengdu, China, in 2018, where he is
currently working toward the Ph.D. degree. His main
research interest focuses on wireless power transfer.

0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.