A DC-Bus Capacitor Discharge Strategy for PMSM Drive System with Large Inertia and Small System Safe Current in EVs

Abstract

When an emergency happens to electric vehicles, the voltage of the DC-bus capacitor which is an important part of the permanent magnet synchronous machine (PMSM) drive system requires to be reduced as quickly as possible. Recently, a new idea of directly using the windings to discharge the capacitor has come forth. This paper proposes a physical energy flow model (EFM) to explain explicitly the winding-based discharge mechanism firstly. In the EFM, the performance characteristics of a classical winding-based discharge schemes are evaluated, but it is found that the discharge time will not be qualified when the machine rotor inertia is large and the system safe current is small. To reject intense voltage surge, a current control algorithm is proposed to bleed the capacitor voltage. Moreover, the proposed current control method shortens the discharge period within 3s for the system studied. The proposed discharge algorithm has been verified in experiment.

Full text

TII-18-3314
 Abstract—When an emergency happens to electric
vehicles, the voltage of the DC-bus capacitor, which is an
important part of the permanent magnet synchronous
machine (PMSM) drive system, requires to be reduced as
quickly as possible. Recently, a new idea of directly using
the windings to discharge the capacitor has come forth.
This paper proposes a physical energy flow model (EFM) to
explain explicitly the winding-based discharge mechanism
firstly. In the EFM, the performance characteristics of the
classical winding-based discharge scheme are evaluated,
but it is found that the discharge time will not be qualified
when the machine rotor inertia is large and the system safe
current is small. In order to reject intense voltage surge, a
current control algorithm is proposed to bleed the
capacitor voltage. Moreover, the proposed current control
method shortens the discharge period to below 3 s for the
system studied. The proposed discharge algorithm is
verified by the experiment which is conducted on a
three-phase PMSM drive system.
Index Terms—Energy flow model, permanent magnet
synchronous machine, winding-based discharge, large
inertia, system safe current.
I. INTRODUCTION
URRENTLY, permanent magnet synchronous machine
(PMSM) drive systems in electric vehicles (EV) are
drawing increasing attention due to their excellent
characteristics of high efficiency and wide speed regulation
range [1]-[10]. Fig.1 shows a commonly used topology of EV
PMSM drive system. In addition to a PMSM, there are four
other main parts, firstly a battery pack providing energy, then a
DC/DC boost converter transferring the low battery voltage
Manuscript received September 07, 2018; revised November 26,
2018; accepted January 18, 2019. (Corresponding author: Yihua Hu).
C. Gong, Y. Hu and K. Ni are with the Department of Electrical
Engineering and Electronics, University of Liverpool, Liverpool L69 3GJ,
U.K. (E-mail: 1452101806@qq.com, y.hu35@liverpool.ac.uk,
k.ni@student.liverpool.ac.uk).
G. Chen is with the School of Aerospace Engineering, Xiamen
University, Xiamen 361005, China. (E-mail: cgp2017@xmu.edu.cn).
H. Wen is with the Department of Electrical and Electronic
Engineering, Xi’an Jiaotong-Liverpool University, Suzhou 215123,
China. (E-mail: Huiqing.Wen@xjtlu.edu.cn).
Z. Wang is with the School of Electrical Engineering, Southeast
University, Nanjing 210096, China. (E-mail: zwang@eee.hku.hk).
into the high voltage that is applied to both the third part, a
DC-bus capacitor absorbing high-frequency power surges, and
the fourth one, a power inverter [11]-[14]. Once emergency
(e.g., car crash) occurs, the system protection device, a circuit
interrupter, will be tripped immediately, and the capacitor
voltage UC is supposed to decrease to a safe level Usafe (60 V)
within 5 s to avoid electrical shock risks according to the United
Nation Vehicle Regulation ECE R94 [15].
Generally, all of the transistors in the three-phase inverter
will be shut off if a fault or failure is detected [16], [17]. If so,
the energy stored in the capacitor cannot be dissipated unless a
bleeder circuit in parallel with the capacitor is adopted [18].
This method can definitely meet the requirement that the
capacitor discharges quickly, but the use of extra circuits
sacrifices the volume, weight and cost of EV drive systems. In
order to achieve a cost-effective voltage discharge solution, a
brand-new idea of directly using the motor windings to
consume the capacitor energy has been presented [11], [19].
This kind of DC-bus capacitor energy consumption method is
characterized by directly controlling the d, q-axis currents in
the PMSM to drive the machine to operate normally in spite of
emergency situations. In this case, the clutch in the vehicle
disconnects the wheels from the traction motor and the motor
just rotates with no load [20]. In [19], the DC-bus capacitor
voltage is dissipated by applying a constant large d-axis current
and zero q-axis current to the machine. But this discharge
algorithm can only be effective when the rotor speed is not
higher than the threshold value that is equal to the speed at
which the back electromotive force (EMF) of the machine is
60V, otherwise the capacitor recharge phenomenon will be
observed because an overcurrent protection mode is triggered
when the bus voltage is close to zero as presented in [11], in
which the entire voltage discharge process is optimized by three
different stages. At the first stage, a nonzero d-axis current and
zero q-axis current are employed to bleed the capacitor voltage.
At the instant that the voltage reaches 60 V, the second stage
launches and a DC-bus voltage regulation algorithm by
applying negative d, q-axis currents to the PMSM is activated
to make the voltage level off until the machine speed is under
the threshold ωmot.th. Finally, both d, q-axis currents will be
ramped down to zero and the residual energy is discharged
completely. The above-mentioned optimization algorithm has
been successfully applied to the high-speed cases when both the
A DC-Bus Capacitor Discharge Strategy for
PMSM Drive System with Large Inertia and
Small System Safe Current in EVs
Chao Gong, Yihua Hu, Senior Member, IEEE, Guipeng Chen, Member, IEEE, Huiqing Wen, Senior
Member, IEEE, Zheng Wang, Senior Member, IEEE, Kai Ni
C

TII-18-3314
DC/DC
Boost
Converter
PMSM
C
Battery
Interrupter
UC
Inverter
Fig. 1. Topology of EV PMSM drive system.
0t0t1t2
5
ωmot.th
Usafe
time (s)
DC-bus voltage Machine speed
Discharge starts
Discharge to 60 VDischarge to 60 V
High/low rotor inertia, large system safe current
High rotor inertia, small system safe current
Fig. 2. Discharge characteristics under different system parameters.
rotor inertia J of the machine and the system safe current Imax
are high without capacitor recharge (the red curve in Fig.2).
However, the DC-bus capacitor voltage characteristics are
closely related to the system parameters when using the
algorithm proposed in [11], especially for the system safe
current that neither causes magnet demagnetization nor exceeds
the maximum allowable current, and the magnitude of rotor
inertia. When J is high but Imax is relatively small (eg., J=0.24
kg·m2 and Imax =100 A), it is difficult to put that optimization
method into practice (the blue curve in Fig.2). For the system
with large inertia and small safe current, the difficulty in
implementing a winding-based discharge algorithm by
controlling the d, q-axis currents is illustrated as follows. The
transient discharge capacity is limited and the motor speed
declines slowly due to small Imax. Consequently, the capacitor
voltage cannot change, which is shown as the red line in Fig.2,
and the discharge time might be over the upper limit. The
easiest way to shorten the discharge time is applying a larger
q-axis but a smaller d-axis current to the motor from the outset,
by which the motor speed drops quickly. However, a larger
negative q-axis current will increase the rate of energy
conversion from the mechanical energy to the electric one
flowing back to the DC-bus side. Therefore, the motor
windings are not able to expend the energy synchronously, and
a huge voltage surge will emerge inevitably because the
capacitance of the DC-bus capacitor is usually not large enough
[21]. In order to avoid excessive voltage surge, it is of great
importance to figure out how to control the q-axis current. Yet
there are no studies about the winding-based discharge control
algorithms for that kind of system that has large rotor inertia
and small system safe current.
This paper proposes a new control method for the PMSM
drive system with large inertia and small system safe current to
quickly discharge the bus capacitor and safely reduce the motor
speed without voltage surge. This method adopts a piecewise
q-axis current locus and makes full use of the maximum
discharge capacity of the system, allowing any active discharge
events regardless of the motor speed. In order to illustrate the
control algorithm at length, this paper presents an energy flow
model (EFM), which is a combination of energetics and
electromagnetics. This model is well suited for capturing the
transient discharge behaviors of a PMSM drive system, and
relying on it, the mechanism and characteristics of both the
traditional and novel winding-based discharge schemes are
discussed.
The structure of the rest of the paper is as follows. Section II
describes an energy flow model (EFM) and presents the
mechanism of the winding-based discharge strategies. Section
III firstly details the characteristics and also the defects of the
published winding-based discharge methods when they are
applied to the system with large inertia and small system safe
current, and then demonstrates the proposed DC-bus voltage
bleeding algorithm. The results of the successful experiments
are in Section IV, and Section V presents the conclusion part.
II. ENERGY FLOW MODEL AND WINDING-BASED
DISCHARGE MECHANISM
Establishing a mathematical or physical model of the drive
system is crucial for detailing the winding-based discharge
mechanism, illustrating the voltage and current properties,
revealing the working status of the motor and designing a novel
discharge algorithm for the system with large inertia and small
system safe current. Thus, before analyzing the winding-based
discharge schemes, the so-called EFM is established.
A. Energy Flow Model
The EFM used for active discharge is shown in Fig.3. The
capacitor is equivalent to a triangular energy tank, and the area
of the shaded triangle represents the electric energy packed into
the tank. The height of the shaded triangle is equal to the
capacitor voltage. s1 and s2 represent two valves whose state
can be controlled by d, q-axis currents. f1 is the energy
consumed by mechanical friction. f2 is the kinetic energy of the
rotor that is going to be converted into electric energy flowing
back to the tank, and most of the energy in the tank (f3) is
expended in the form of winding heat loss.
Assume that the active discharge is requested at t0 when the
capacitor voltage is Ufull. The initial energy Q0 stored in the tank
is as follows:
2
01
2full
Q C U
(1)
where C is the capacitance.
After a period of △t, the energy flow can be expressed as:
2
1m
t
f F dt




(2)
__
22
21
1()
2iq st iq ed
f J f

   
(3)
22
3d s q s
tt
f i R dt i R dt

   

(4)

TII-18-3314
Usafe
Uc
Ufull
C·Ufull
C·Uc
Energy
Tank
s2
f3
f1
f2
Mechanical
Energy
s1
iq 0, s1 is on.
Frictional Loss Winding Loss
PMSM
iq 0
or
id 0
if
if , s2 is on.
iq = 0
and
id = 0
if , s1,2 is off.
Fig. 3. Energy flow model used for active discharge.
where ωiq_st and ωiq_ed are the rotor mechanical angular speeds
before and after iq is injected to the machine, respectively. That
is, only when iq is negative could f2 exist. The stator winding
resistance is Rs, and F is the viscous coefficient. ωm represents
the instantaneous rotor mechanical angular speed. Then, the
remaining energy Qrem in the capacitor is
0 2 3rem
Q Q f f  
(5)
The capacitor voltage Uc at the end is
2rem
cQ
UC


(6)
The above is the energetic part of the proposed EFM, which
is based on the law of the conservation of energy. However, that
is not enough for accurately evaluating the transient
performance characteristics of the drive system and calculating
the amount of energy conversion. To complement the
deficiency, an electromagnetic part is added to the EFM
[22]-[25]:
q
d s d
d m q
d d d
L
di R u
i p i
dt L L L

   
(7)
q q f
ds
m d q m
q q q q
di u
LR
p i i p
dt L L L L


    
(8)
1(1.5 ( ( ) ) )
mf q d q d q m
dp i L L i i F
dt J

    
(9)
where ud, uq are d, q-axis control voltages. Ld, Lq are d, q-axis
inductances. Additionally, p represents the number of machine
pole pairs and ψf is the permanent magnet flux linkage.
B. Winding-based Discharge Mechanism
The winding-based discharge mechanism includes, firstly,
the winding resistance, which functions as a bleeder load.
Specifically, an external emergency does not trigger the
protection regime immediately. Instead, an active discharge
operating mode by using the windings to consume the energy
stored in the capacitor launches prior to it. Although that
process is able to bring down the capacitor voltage, it is highly
required to shorten the discharge period as much as possible (5
s is long in this aspect), so as to shut down the entire drive
system in time. Besides, according to the different states of s1
and s2 (directly related to id and iq) in Fig.3, the discharge
mechanism can be expressed as four basic cases shown in Table
I. a) When both iq and id are zero, s1 and s2 are off. There is only
one path to consume the rotor kinetic energy by f1, while the
electric energy in the tank cannot flow out. Namely, f2=f3=0.
The capacitor voltage remains at the maximum level Ufull. In
this case, the discharge time will be very long since the PMSM
may be freely spinning (multiple minutes in duration).
b) When iq is kept at zero and id is negative, s1 is off but s2 is
on. The rotor mechanical energy cannot be converted into
electric energy, and it is consumed only by friction, but the
energy stored in the tank can be rapidly radiated because of the
existence of windings (f2=0, f3≠0).
c) When iq is nonzero and id is zero, both s1 and s2 are on.
Most of the rotor mechanical energy will be transformed into
capacitor electric energy, and then consumed by the windings
together with the initial energy in the tank (f2≠0, f3≠0). At the
same time, a small percentage of kinetic energy is consumed by
the path of f1.
d) When negative iq and id are injected into the machine, s1
and s2 are in open states. A large amount of rotor mechanical
energy floods into the tank (f2≠0, f3≠0) but the capacitor energy
is dissipated by f3 very quickly.
A classic winding-based discharge process always contains
one or more of the aforementioned cases. In [19], only the case
b) is tried out to achieve the bus capacitor voltage discharge.
From the very beginning to the end, the d-axis reference current
is set as the safe current (Imax) of system, and the q-axis
reference current is zero. In [11], the discharge process is
divided into three phases, the first of which employs case b),
and both the second and the third of which adopt case d) by
applying nonzero d, q-axis currents to the machine. The
proposed method in this paper only utilizes the fourth case d),
which will be detailed in Section III.
III. WINDING-BASED DISCHARGE APPROACHES APPLIED
TO THE SYSTEM WITH LARGE INERTIA AND SMALL SAFE
CURRENT
For the purpose of redesigning an effective and
high-efficiency discharge approach, it is significant to discover
beforehand the characteristics and defects of the traditional
winding-based bleeding schemes. In virtue of the EFM, the
diverse characteristics of different discharge methods are
analyzed and a novel algorithm based on direct current control
is presented in this section.
In order to implement an intuitive discussion on the
winding-based discharge strategies, a PMSM drive system for
TABLE I
MECHANISM OF WINDING-BASED DISCHARGE SCHEMES
Current states
Valve status
Proportion of energy flow
id
iq
s1
s2
f1
f2
f3
zero
zero
off
off
large
zero
zero
nonzero
zero
off
on
large
zero
nonzero
zero
nonzero
on
on
small
nonzero
nonzero
nonzero
nonzero
on
on
small
nonzero
nonzero

TII-18-3314
0
150
300
450
1 2 3 4 5
Capacitor voltage
(V)
172V
492V
Usafe
3.2
600
(b)
Time (s)
140rad/s
Angular speed
(rad/s)
0
100
200
300
400
ωm
Uc
0
100
200
300
Capacitor voltage
(V)
(a)
1 2 3 4 5
Usafe
170V
Time (s)
Angular speed
(rad/s)
0
100
200
300
400
400
98 rad/s
ωm
Uc
recharge
slow discharge (generator running state)
rapid discharge
slow discharge (generator running state)
Fig. 4. Capacitor voltage and speed characteristics of the traditional
winding-based algorithms. (a) Traditional method a. (b) Traditional
method b.
-120
-80
-40
0
d-axis current (A)
q-axis current (A)
Time (s)
-40
-20
0
1 2 345
0
Imax
passive change
active change
∆iq=10 A
12traditional method a
traditional method b
1desired d-axis current 2declining d-axis current
traditional method b
traditional method a
Fig. 5. d,q-axis current characteristics of the traditional winding-based
algorithms.
EV with parameters in Table II is studied. The active discharge
starts when the motor speed is at the rated value ωrated.
A. Discharge Characteristics and Defects of Traditional
Winding-based Discharge Schemes
When the bleeding algorithm in [19] (traditional method a) is
applied to the system with large inertia and small safe current,
the initial DC-bus voltage is promptly reduced. But it cannot
drop under the safe level as quickly as expected so that the
capacitor is charged when the back EMF is higher than the
DC-bus capacitor voltage, slowing the discharge rate. The
simulation results of capacitor voltage and speed characteristics
in Fig.4 (a) proved the above statement. The discharge is
requested at 0.5 s when -100 A d-axis and zero q-axis currents
are utilized. The capacitor voltage immediately jumped down
to 170 V, after which it declines slowly and reaches the safe
value until 5.5 s (the discharge time is 5.0 s) when the speed
reaches 98 rad/s. The reason why the capacitor voltage can
only decrease to 170 V in the rapid discharge process, instead
of decreasing to under 60 V as desired is that the highest
allowable current of this system is so low that the instantaneous
discharge capacity is limited. In detail, when the negative id
(-100 A) is applied, the air-gap field of the motor is weakened
instantly, reducing the back EMF simultaneously to about 170
V at the rated speed. At the moment, the capacitor voltage is
much higher than the back EMF, so that the energy can flow in
one direction from the capacitor to PMSM. But soon the
capacitor voltage will not be constantly higher than the back
EMF due to the gradually declining rotor speed. Afterwards,
the capacitor will be in a state of discharge accompanied by
charge (slow discharge), and only when the rotor speed goes
down to the position where the back EMF is equal to Usafe will
the capacitor voltage remain under 60 V.
The above-mentioned voltage and speed characteristics of
the traditional method a also indicate that the first stage of the
algorithm in [11] cannot be achieved, not to mention then
stabilizing the voltage at a value slightly under Usafe. Hence, it is
impracticable to directly apply that current control algorithm to
the system with large inertia and small safe current. But in order
to shorten the discharge time in Fig.4 (a), we can learn from the
second stage of the discharge algorithm in [11], solving the
problem by controlling the q-axis current to keep at a higher
level (traditional method b) so as to speed up the deceleration
process. Fig.4 (b) illustrates the simulation results of the
process. In this situation, the d-axis of (-98 A) and q-axis
current of (-20 A) are adopted as the reference values at 0.5 s.
Compared to Fig.4 (a), the DC-bus voltage reaches 60 V more
quickly within 3.2 s (the discharge time is 2.7 s) when the speed
arrives at about 140 rad/s. However, there is an obvious voltage
surge in a short period after the beginning of discharge, and if
the reference value of iq is increased, the peak of the voltage
surge will get higher. This happens because a larger negative
q-axis current can increase the energy conversion rate from the
mechanical energy to electric energy, which is more significant
when the machine speed is high, but the windings are not able
to expend the energy synchronously because the total discharge
current does not increase.
The current characteristics of the traditional winding-based
algorithms are compared in Fig.5 where two interesting
phenomena can be seen. Firstly, the q-axis current in the
windings turns to be negative passively without following the
planned trajectory for the method a, which seems to correspond
to case d), but iq is actively generated in the system rather than
being passively injected as in the method b. Secondly, the
discharge current, especially the d-axis current, will witness an
downward trend when the DC-bus voltage is not large enough
to maintain it at the reference level for both methods. At the
time, the energy consumption rate by the machine windings
TABLE II
PMSM DRIVE SYSTEM PARAMETERS
Parameter
Value
Unit
stator winding resistance Rs
0.275
Ω
d-axis inductance Ld
0.8
mH
q-axis inductance Lq
0.8
mH
the number of pole pairs p
3
-
moment of inertia J
0.24
kg·m2
viscous coefficient F
0.0035
-
permanent magnet flux linkage Ψf
0.18
Wb
DC-bus voltage VDC-bus
310
V
system safe current Imax
100
A
rated speed ωrated
345
rad/s
DC-bus capacitor C
560
μF

TII-18-3314
time (s)
t0
0
∆t∆t
t1t2
ω0=ωrated
Angular speed
ωm (rad/s)
ti
ti-1
ω1 ω2
∆t ωi
iq_ref1
q-axis current
iq_ref (A)
ωi-1 ωave_i
ωave_2
ωave_1
iq_ref2
iq_refi
Fig. 6. q-axis reference current and desired speed characteristics in the
entire discharge process.
goes down at the same time. Therefore, another effective way
to shorten the discharge time is to extend the duration of
applying a larger discharge current. Apart from faster
deceleration, it is also in that way for the method b to achieve a
shorter discharge period. In detail, firstly, even though id falls
just by about 2 A (2%), it lasts 0.8 s longer at the desired
position before declining. Secondly, iq can remain at nearly -20
A during the whole process, and only when the bus voltage
reduces almost to zero will it decline in method b.
Comparatively, the q-axis current generated in method a is
lower (△iq) and experiences a declining trend from 2.5 s.
In terms of the defects, the discharge time for the traditional
method a is not qualified. Although the traditional method b
shows more attractive characteristic from this aspect, namely
shorter discharge time, the voltage recovery is unwanted for the
sake of system safety. As a consequence, Section III-B will
explain a novel DC-bus capacitor discharge strategy by
calculating and applying appropriate q-axis reference currents
to shorten the discharge period, and meanwhile avoiding
voltage rise effectively.
B. Proposed Discharge Algorithm
By comparing the bleeding properties of the above two
algorithms, it is feasible to shorten the overall bleeding time by
using a relatively high q-axis reference current iq_ref. However,
it is also found that when the q-axis current in the machine is
about -10 A from the beginning, there is no voltage rise
phenomenon. But when it is controlled at -20 A, the voltage
surge shows up in the high speed range. So the locus of iq_ref
should be well designed if a high q-axis current control strategy
is employed to shorten the discharge time as well as eliminate
the voltage surge. Now, a new issue of how to determine the
reference signal arises.
When the negative reference iq_ref and d-axis reference
current id_ref keep fixed, a constant electromagnetic braking
torque Te will be produced once the currents in the PMSM are
controlled to track the reference values.
_ _ _
1.5 ( ( ) )
1.5 ( ( ) )
e f q d q d q
f q ref d q d ref q ref
T p i L L i i
p i L L i i
   
   
(10)
Ignoring the term of reluctance torque for simplicity, Te can be
expressed as:
_
1.5
e f q ref
T p i
(11)
So the motor will slow down with a constant deceleration adec.
_
||1.5
e
dec f q ref
T
a p i
JJ
  
(12)
After a period of △t, the kinetic energy converted into the
electric energy f2 in (3) can be rewritten as:
_
22
2 _ _ 1
9
1.5 ( )
8
iq st
f q ref f q ref
f p i t p i t f
J

      
(13)
Therefore, f2 is proportional to the machine speed if a constant
q-axis current is adopted. Thus, the case that the energy
conversion rate (from mechanical energy to electric energy) is
high but the windings cannot consume the energy
synchronously (f2 > f3) happens only in the high speed range.
The problem can be solved by employing piecewise q-axis
current as in Fig.6. The entire discharge process is divided into
several periods of △t. iq_ref is smaller when the speed is high,
and gets larger and larger over time. On the basis of EFM, the
key technique proposed in this paper to avoid the voltage surge
can be described as letting f2 ≤ f3 over each period of △t. In
other words, the phenomenon will vanish whenever the
mechanical energy conversion rate is not higher than the energy
consumption rate. The computing method for iq_refi is shown
below.
In this paper, we try to ensure the maximum capacity of
discharge, so the system current is expected to be controlled to
maintain at Imax during the whole process, that is,
2 2 2
_ _ md ref q ref ax
i i I
(14)
Take the ith interval as an example, on the basis of the
formulas (4) and (10),
2
3max s
f I R t  
(15)
f2 is also equal to the work done by the braking torque, namely,
2||
em
t
f T dt




(16)
When △t is short, the angular speed ωm can be approximated as
the average value ωave_i from ti-1 to ti. So f2 can be rewritten as:
2_
||
e ave i
f T t

  
(17)
1
_2
ii
ave i




(18)
ω0 is the initial speed when discharge is requested. In order to
calculate ωi, assume that the mechanical friction torque is much
smaller than Te and ignore its influence on the deceleration. ωi
can be expressed as:
1i i dec
at


  
(19)
According to the requirement f2 ≤ f3 and (12), iq_refi can be
calculated then, that is:
22
1 1 m
_
2
1.5
i i ax s
q ref i f
tIR
J
ipt
J



  

(20)
Substitute (20) into (14), a piecewise d-axis reference current
locus can be derived as:
22
_ m _d refi ax q refi
i I i
(21)
For the purpose of reducing the motor speed as quickly as
possible, the absolute value of iq_refi needs to be maximum

TII-18-3314
PWM
signal
iq
PI
controller
id
αβ
dq
22
1 1 m
2
1.5
i i ax s
f
tIR
J
pt
J



  

PI
controller
id_ref = sqrt(Imax2-iq_ref2)
iq_ref
Reference locus
Modulation
Current regulation
Fig. 7. Block diagram of the proposed discharge.
0
100
200
300
-120
-80
-40
0
-40
-20
0
Capacitor voltage
(V)
(a)
1 2 34 5
d-axis current
(A)
q-axis current
(A)
(b)
1 2 345
0
Usafe
Time (s)
Time (s)
Angular speed
(rad/s)
0
100
200
300
400
400
3.7
ωm
Uc
110 rad/s
iq_ref
iq
-10.6
-11.3
-12.1
-13.1-14.4 -16.2 -23.4 -35.4
-18
-99.4 -99.3 -99.2 -99.1-98.9-98.6-98.3-97.2 -93.5
id_ref
id
Fig. 8. Characteristics of the proposed winding-based algorithm. (a)
Capacitor voltage and PMSM mechanical angular speed. (b) d,q-axis
current.
(a)
C
Converter
Udc
1
2
Signal detection
Signal processing
q-axis current
controller
d-axis current
controller
Drive signal
generation
1
2
t
speed controller
1
2
sqrt(Imax2-iq_ref2)
0
id_ref
iq_ref
ω0
ωm
ia
ib
ic
θ: rotor position
UC
0-20
ab
-Imax -98
ab
θ
a: traditional method a
b: traditional method b
(b)
Fig. 9. Experimental system. (a) Photo of the experimental setup.
(b) Schematic diagram.
within the permissible range. Fig.7 illustrates the block diagram
of the discharge method. Two current controllers and a pulse
width modulation (PWM) signal generator are adopted in the
topology for discharge.
Besides, it should be acknowledged that the shorter △t is, the
more representative and typical ωave_i will be. Assume that the
change of the machine parameters can be ignored during
discharge process and when △t is set as 0.5 s and ω0=ωrated,
Fig.8 shows the values of id_ref and iq_ref for the different
intervals and the corresponding simulation results. In
comparison with the method a, firstly, the capacitor voltage
falls to 60 V within 3.7 s (discharge time is 3.2 s) when the
speed gets to about 110 rad/s for the novel method, which
becomes shorter and satisfies the discharge requirement. The
reason why the threshold speed for the novel method is higher
than that for traditional method a (but lower than traditional
method b) is that the back-EMF of the machine is affected by
the flux-weakening d-axis current. The larger the negative id is,
the higher level the rotor speed should stand at for obtaining the
same voltage. Secondly, the q-axis current in the motor can
track the reference values well and the d-axis current starts to
decline from about 2.5 s, proving that the new current control
algorithm is capable of extending the duration of applying
relatively large discharge currents to the machine. Then, the
voltage surge phenomenon is completely avoided, contributing
to eliminating the electrical shock risks of the EV powertrains.
Finally, an interesting phenomenon is that the current
references are derived by setting f2=f3 but the bus voltage
continues to drop during discharge. This happens because
firstly, the energy consumed by friction (f1) is ignored when
designing the current reference locus. Considering that the
mechanical friction can expend energy, the capacitor voltage
will decline. Secondly, it can be noticed that at the start of each
period in Fig.8 (b), there exist current overshoots of the real iq.
This will cause that the speed declines faster over the former
half period of △t than that of the latter half period.
Consequently, the calculated f2 is smaller than the real value,
which will in turn lead to that the calculated q-axis reference
current is relatively lower. Thus, although we expect that f2 = f3,
f2 is less than f3 in reality. But the goal to avoid voltage surge
can be achieved. The characteristics of the proposed
winding-based method will also be verified by experiments in
Section IV.
IV. EXPERIMENTAL RESULTS
Experiments are conducted on a three-phase PMSM whose
parameters are also consistent with Table II. The experimental
equipment is shown in Fig.9 (a). A DC power supply is
available at 310 V. An intelligent power module (IPM),
Mitsubishi PM100RLA120, is used as the voltage source
inverter with the frequency of 10 kHz. Four thin-film capacitors,
DHF DAWNCAP 140 μF, are connected in parallel to compose
the desired 560 μF DC-bus capacitor. The proposed discharge
algorithm is implemented on DSP TMS320F28335 controller
board. The real rotor position is detected by a rotary

TII-18-3314
time [0.5s /div]
(a)
(c) (d)
Usafe
time [1.5 ms/div]
(b)
ib
iaic
Uc [192.5V/div]
ωm [250rad/s/div]
current [40A/div]
id [50A/div]
time [0.5 s/div]
iq [25A/div] iqref = -20A
f2 [5000J/div]
f3 [5000J/div]
time [0.5s /div]
460V
idref = -98A
Fig. 11. Experimental results of the traditional discharge method b when
emergency occurs at the speed of 345 rad/s. (a) DC-bus capacitor
voltage and machine speed. (b) Phase current. (c) d,q-axis current. (d)
Amount of energy conversion and consumption by windings.
time [0.5s /div]
(a)
(c) (d)
Usafe
time [1.5 ms/div]
(b)
ib
iaic
Uc [192.5V/div]
ωm [250rad/s/div]
current [40A/div]
id [50A/div] 1
time [0.5 s/div]
iq [25A/div] iqref
1cross-co upling effect
f2 [5000 J/di v]
f3 [5000 J/di v]
time [0.5s /div]
Fig. 12. Experimental results of the proposed discharge algorithm when
emergency occurs at the speed of 345 rad/s. (a) DC-bus capacitor
voltage and machine speed. (b) Phase current. (c) d,q-axis current. (d)
Amount of energy conversion and consumption by windings.
time [0.3 s/div]
(a)
(c) (d)
(b)
ωm [125rad/s/div]
-18.5 -33.2
-22.8
Usafe Uc [192.5V/div]
ib
ia
ic current [40A/div]
time [2.5 ms/div]
1
1cross-co upling effect id [50A/div]
iq [25A/div]
time [0.3s /div]
iqref
f2 [2500J/div]
f3 [2500J/div]
time [0.3 s/div]
Fig. 13. Experimental results of the proposed discharge algorithm when
emergency occurs at the speed of 200 rad/s. (a) DC-bus capacitor
voltage and machine speed. (b) Phase current. (c) d,q-axis current. (d)
Amount of energy conversion and consumption by windings.
time [0.6 s/div]
(a)
(c) (d)
time [1.5 ms/div]
(b)
ib
ia
ic
ωm [250rad/s/div]
current [40A/div]
id [50A/div]
time [0.6 s/div]
iq [25A/div] iqref = 0
f2 [5000 J/div]
f3 [5000 J/div]
time [0.6 s/div]
idref = -100A
Usafe Uc [192.5V/div]
Fig. 10. Experimental results of the traditional discharge method a when
emergency occurs at the speed of 345 rad/s. (a) DC-bus capacitor
voltage and machine speed. (b) Phase current. (c) d,q-axis current. (d)
Amount of energy conversion and consumption by windings.
transformer. Hall current sensors, HNC-100LT, are used to
measure three-phase currents while the motor d, q-axis currents
are calculated by the digital controller. The DC-bus voltage is
measured by a voltage transducer LV25-P. Using the USB-
RS485 communication interface, the collected data are
transmitted to and further recorded by the host computer. The
schematic diagram of the traditional and proposed discharge
methods is presented in Fig.9 (b). A programmatic virtual
switch is used to select the operating state of the system. When
the system works normally, port 2 is connected and the system
can be controlled by any PMSM drive technique, such as
double closed-loop speed regulation strategy [26] and direct
torque control method [27]. Once an emergency occurs, port 1
is connected and the winding-based discharge algorithm is
implemented.
Assume the active discharge request occurs at 0.5 s when the
motor speed is ωrated. Fig.10 (a) illustrates that the capacitor
voltage of the traditional method a drops to the safe level at
nearly 5.75 s, being a little bit longer than that in the simulation
result. Fig.10 (b) and (c) show the characteristics of the three
phase currents from 0.52 to 0.5215s and the d, q-axis currents,
respectively. Before 2.4 s, the d-axis current can track the
prescribed trajectory well, after which it goes down gradually.
In accordance with the simulation results, the q-axis current
(about 14 A) is generated automatically at the time when the
discharge is requested, while it starts to decline from 2.5 s and
reaches below -5 A at 6.0 s. Additionally, f2 and f3 can be
calculated according to the EFM, and Fig.10 (d) demonstrates
that about 14000 J of mechanical energy is converted into the
electric energy and then consumed by the windings by the end
of discharge.
Fig.11 depicts the discharge characteristics of the traditional
discharge method b. The discharge time is 2.8 s, which is very
close to that in the simulation result in Fig.4 (b). However, a
marked voltage surge (about 460 V) appears in the high speed
range. In terms of the current characteristics, iq can stay at the
desired value during the whole process while id witnesses an
downward trend from 3.1 s. Similar to method a, a total of
approximately 14000 J of energy is expended by the windings
during the whole process.
Fig.12 demonstrates the experimental results of the proposed
bleeding algorithm. Overall, it can be noticed that the DC-bus
capacitor voltage drops to 60 V at nearly 3.5 s and the voltage
surge disappears. The reason why the proposed algorithm can
avoid remarkable voltage surge in the PMSM drive system with
large inertia is that a piecewise q-axis current locus is used. The
higher the mechanical speed is, the relatively lower iq is applied
to the machine, as shown in Fig.12 (c), producing smaller
braking electromagnetic torque. Consequently, the motor speed

TII-18-3314
declines more slowly in the high speed range, making it
possible that the energy consumption rate by windings is no
lower than the mechanical energy conversion rate. That stands
in contrast with Fig.11 (c) in which iq always keeps at a high
position and the motor windings are incapable of consuming
the converted mechanical energy synchronously. Before
leaving Fig.12 (c), we find that the d-axis current begins to fall
from 2.8 s, proving the statement that the new algorithm can
extend the duration of applying relatively large discharge
current to the motor compared to method a.
Table III shows the features of all the three above-mentioned
bus voltage bleeding algorithms when they are applied to a
drive system with large rotor inertia and small safe current. As
far as the proposed method is concerned, although the
discharge time is about 3.0 s, which is a little bit longer than 2.8
s for method b, it has been shortened greatly by 42.8%
compared to that in the traditional strategy a. More importantly,
the large voltage surge disappears. This is crucial to strengthen
the security of both the EV components and passengers,
especially in an emergency.
In order to verify that the proposed algorithm is effective
regardless of the machine speed, as shown in Fig.13, the
experiment was carried out when assuming an active discharge
is requested at the speed of 200 rad/s. For the sake of low initial
speed, the calculated q-axis reference current gets lower (-18.5
A) at first. The DC-bus voltage descends to the safe level at
about 1.35s and no large voltage recovery is witnessed. Only
about 4500 J of kinetic energy is transferred to electric energy
and consumed by the machine windings when the rotor speed
arrives at zero.
Looking at the current properties in Fig.12 (c) and Fig.13 (c)
in detail, the d-axis current fluctuates at the moment when a
larger q-axis current is applied to the machine in the proposed
algorithm. This phenomenon arises from the cross-coupling
effect [28]-[30]. Besides, the q-axis current can level off at the
desired value until the DC-bus voltage is nearly zero, while the
d-axis current starts to decline in advance in accord with the
simulation results. That indicates that the d-axis current is more
sensitive to the declining DC-bus voltage in the machine when
a winding-based bleeding algorithm is employed. Therefore, a
nonzero q-axis current control algorithm is more suitable for
bleeding the DC-bus voltage.
V. CONCLUSION
It is cost-effective to directly use PMSM windings rather
than external bleeder circuits for discharging the DC-bus
capacitor energy, and several winding-based discharge
methods emerged. However, they are not effective when the
drive system has large rotor inertia and small safe current.
In the first place, this paper proposes an EFM model, which
is a combination of energetics and electromagnetics, and it is
utilized to illustrate the mechanism of the traditional
winding-based discharge algorithms and performance
characteristics in the process of bleeding. Secondly, aiming at
the system with large rotor inertia and low safe current, this
research analyzes the defects of the traditional methods and
proposes a segmented nonzero d, q-axis current control
algorithm, avoiding large voltage surge and shortening the
discharge time simultaneously. The simulation and
experimental results prove that the proposed current control
algorithm is more suitable for bleeding the capacitor voltage of
the PMSM drive systems with large inertia and small system
safe current in EVs compared to the traditional methods. In
conclusion, the proposed winding-based discharge approach
can avoid voltage breakdown hazard if an emergency occurs,
improving the safety of the EV traction systems regardless of
the system parameters, especially the maximum allowable
current and rotor inertia of the traction PMSM.
REFERENCES
[1] M. Chen and J. Lu, “High-Precision Motion Control for a Linear
Permanent Magnet Iron Core Synchronous Motor Drive in Position
Platform, ” in IEEE Transactions on Industrial Informatics, vol. 10, no. 1,
pp. 99-108, Feb. 2014.
[2] J. Lara, J. Xu and A. Chandra, “Effects of Rotor Position Error in the
Performance of Field-Oriented-Controlled PMSM Drives for Electric
Vehicle Traction Applications,” in IEEE Transactions on Industrial
Electronics, vol. 63, no. 8, pp. 4738-4751, Aug. 2016.
[3] J. Lu, X. Zhang, Y. Hu, J. Liu, C. Gan and Z. Wang, “Independent Phase
Current Reconstruction Strategy for IPMSM Sensorless Control Without
Using Null Switching States,” in IEEE Transactions on Industrial
Electronics, vol. 65, no. 6, pp. 4492-4502, June 2018.
[4] P. L. Xu and Z. Q. Zhu, “Novel Carrier Signal Injection Method Using
Zero-Sequence Voltage for Sensorless Control of PMSM Drives,” in
IEEE Transactions on Industrial Electronics, vol. 63, no. 4, pp.
2053-2061, April 2016.
[5] R. Ni, D. Xu, F. Blaabjerg, K. Lu, G. Wang and G. Zhang, “Square-Wave
Voltage Injection Algorithm for PMSM Position Sensorless Control With
High Robustness to Voltage Errors,” in IEEE Transactions on Power
Electronics, vol. 32, no. 7, pp. 5425-5437, July 2017.
[6] G. Feng, C. Lai, K. L. V. Iyer and N. C. Kar, “Improved High-Frequency
Voltage Injection Based Permanent Magnet Temperature Estimation for
PMSM Condition Monitoring for EV Applications,” in IEEE
Transactions on Vehicular Technology, vol. 67, no. 1, pp. 216-225, Jan.
2018.
[7] Y. Mao, J. Yang, T. Wang, D. Yin and Y. Chen, “High dynamic
sensorless control for PMSMs based on decoupling adaptive observer,”
2016 IEEE Energy Conversion Congress and Exposition, Milwaukee,
WI, 2016, pp. 1-8.
[8] T. Wang, C. Liu, G. Lei, Y. Guo and J. Zhu, “Model predictive direct
torque control of permanent magnet synchronous motors with extended
set of voltage space vectors,” in IET Electric Power Applications, vol. 11,
no. 8, pp. 1376-1382, Sep. 2017.
[9] H. Zhu, X. Xiao and Y. Li, “Torque Ripple Reduction of the Torque
Predictive Control Scheme for Permanent-Magnet Synchronous Motors,”
in IEEE Transactions on Industrial Electronics, vol. 59, no. 2, pp.
871-877, Feb. 2012.
[10] F. Mwasilu and J. W. Jung, “Enhanced Fault-Tolerant Control of Interior
PMSMs Based on an Adaptive EKF for EV Traction Applications,” in
IEEE Transactions on Power Electronics, vol. 31, no. 8, pp. 5746-5758,
Aug. 2016.
[11] Z. Ke, J. Zhang and M. W. Degner, “DC Bus Capacitor Discharge of
Permanent-Magnet Synchronous Machine Drive Systems for Hybrid
Electric Vehicles,” in IEEE Transactions on Industry Applications, vol.
53, no. 2, pp. 1399-1405, March-April 2017.
[12] J. Zheng, Z. Wang, D. Wang, Y. Li and M. Li, “Review of fault diagnosis
of PMSM drive system in electric vehicles,” 2017 36th Chinese Control
Conference, Dalian, 2017, pp. 7426-7432.
[13] P. Dost and C. Sourkounis, “On Influence of Non Deterministic
Modulation Schemes on a Drive Train System With a PMSM Within an
Electric Vehicle,” in IEEE Transactions on Industry Applications, vol.
52, no. 4, pp. 3388-3397, July-Aug. 2016.
TABLE III
FEATURES OF DC-BUS VOLTAGE BLEEDING ALGORITHMS
Type of algorithm
Discharge time (s)
Voltage surge (V)
Traditional method a
5.25
none
Traditional method b
2.8
460
Proposed method
3.0
none

TII-18-3314
[14] A. Choudhury, P. Pillay and S. S. Williamson, “DC-Link Voltage
Balancing for a Three-Level Electric Vehicle Traction Inverter Using an
Innovative Switching Sequence Control Scheme,” in IEEE Journal of
Emerging and Selected Topics in Power Electronics, vol. 2, no. 2, pp.
296-307, June 2014.
[15] United Nation Economic Commission for Europe Vehicle Regulation,
No.94 (ECE R94), Unifrom provisions concerning the approval of
vehicles with regard to the protection of the occupants in the event of a
frontal collision, Rev. 2, Annex 11, Aug. 2013.
[16] Chao Gong, Jinglin Liu, Zexiu Han and Haozheng Yu, “Analysis on
uncontrolled generation in electrical vehicles and a battery protection
method,” 2017 IEEE 3rd International Future Energy Electronics
Conference and ECCE Asia, Kaohsiung, 2017, pp. 1606-1610.
[17] Y. Zhang and T. M. Jahns, “Uncontrolled Generator Operation of PM
Synchronous Machine Drive With Current-Source Inverter Using
Normally on Switches,” in IEEE Transactions on Industry Applications,
vol. 53, no. 1, pp. 203-211, Jan.-Feb. 2017.
[18] A. K. Kaviani, B. Hadley and B. Mirafzal, “A Time-Coordination
Approach for Regenerative Energy Saving in Multiaxis Motor-Drive
Systems,” in IEEE Transactions on Power Electronics, vol. 27, no. 2, pp.
931-941, Feb. 2012.
[19] T. Goldammer, T. Le, J. Miller, and J. Wai, “Active high voltage bus
bleed down,” U.S. Patent 20120161679 A1, Jun. 28, 2012.
[20] T S. Ashida, K. Yamada, M. Nakamura, T. Shimana, and T. Soma,
“Electric vehicle, and control apparatus and control method for electric
vehicle,” U.S.Patent 8631894 B2, Jan. 21, 2014.
[21] G. J. Su and L. Tang, “A segmented traction drive system with a small dc
bus capacitor,” 2012 IEEE Energy Conversion Congress and Exposition,
Raleigh, NC, 2012, pp. 2847-2853.
[22] Z. Mynar, L. Vesely and P. Vaclavek, “PMSM Model Predictive Control
With Field-Weakening Implementation,” in IEEE Transactions on
Industrial Electronics, vol. 63, no. 8, pp. 5156-5166, Aug. 2016.
[23] T. Tarczewski and L. M. Grzesiak, “Constrained State Feedback Speed
Control of PMSM Based on Model Predictive Approach,” in IEEE
Transactions on Industrial Electronics, vol. 63, no. 6, pp. 3867-3875,
June 2016.
[24] J. Linares-Flores, C. García-Rodríguez, H. Sira-Ramírez and O. D.
Ramírez-Cárdenas, “Robust Backstepping Tracking Controller for
Low-Speed PMSM Positioning System: Design, Analysis, and
Implementation,” in IEEE Transactions on Industrial Informatics, vol.
11, no. 5, pp. 1130-1141, Oct. 2015.
[25] G. Feng, C. Lai and N. C. Kar, “Particle-Filter-Based Magnet Flux
Linkage Estimation for PMSM Magnet Condition Monitoring Using
Harmonics in Machine Speed,” in IEEE Transactions on Industrial
Informatics, vol. 13, no. 3, pp. 1280-1290, June 2017.
[26] L. Sun, X. Wang, Z. Deng and Q. Ding, “Research on stable regions of
double-closed loop system for PMSM with low-pass filters, ” 2014 17th
International Conference on Electrical Machines and Systems,
Hangzhou, 2014, pp. 1150-1156.
[27] F. Niu, B. Wang, A. S. Babel, K. Li and E. G. Strangas, “Comparative
Evaluation of Direct Torque Control Strategies for Permanent Magnet
Synchronous Machines,” in IEEE Transactions on Power Electronics,
vol. 31, no. 2, pp. 1408-1424, Feb. 2016.
[28] A. Rabiei, T. Thiringer, M. Alatalo and E. A. Grunditz, “Improved
Maximum-Torque-Per-Ampere Algorithm Accounting for Core
Saturation, Cross-Coupling Effect, and Temperature for a PMSM
Intended for Vehicular Applications,” in IEEE Transactions on
Transportation Electrification, vol. 2, no. 2, pp. 150-159, June 2016.
[29] Xiaochun Fang, Fei Lin and Zhongping Yang, “A modified
flux-weakening control method of PMSM based on the d-q current
cross-coupling effect,” 2014 IEEE Conference and Expo Transportation
Electrification Asia-Pacific, Beijing, 2014, pp. 1-6.
[30] H. Ge, B. Bilgin and A. Emadi, “Global loss minimization control of
PMSM considering cross-coupling and saturation,” 2015 IEEE Energy
Conversion Congress and Exposition, Montreal, QC, 2015, pp.
6139-6144.
Chao Gong was born in Shandong province in
P.R. China, on February 22, 1991. He received
the B.Eng. and the M.Eng degree in electrical
engineering from Northwestern Polytechnical
University, Xi’an, China, in 2014 and 2016,
respectively. Currently, he is a PHD student in
the University of Liverpool.
His research interests include electrical
machines design and drives, power electronics
and motion control.
Yihua Hu (M’13-SM’15) received the B.S.
degree in electrical motor drives in 2003, and the
Ph.D. degree in power electronics and drives in
2011. Between 2011 and 2013, he was with the
College of Electrical Engineering, Zhejiang
University as a Postdoctoral Fellow. Between
2013 and 2015, he worked as a Research
Associate at the power electronics and motor
drive group, the University of Strathclyde.
Currently, he is a Lecturer at the Department of
Electrical Engineering and Electronics, University of Liverpool (UoL). He
has published 65 papers in IEEE Transactions journals. His research
interests include renewable generation, power electronics converters &
control, electric vehicle, more electric ship/aircraft, smart energy system
and non-destructive test technology. He is the associate editor of IET
Renewable Power Generation, IET Intelligent Transport Systems and
Power Electronics and Drives.
Guipeng Chen (M’18) received the B.E.E.
degree in electrical engineering from Zhejiang
University, Hangzhou, China, in 2011, and the
Ph.D. degree in power electronics and electric
drives from the College of Electrical Engineering,
Zhejiang University, in 2017. During the PHD
study, he joined Fuji Electric Matsumoto Factory
as a summer intern in 2014 and was invited to the
University of Liverpool as a research assistant for
a half-year program from July 2016. He is
currently working as a Postdoc at the Instrument Science and
Technology Postdoc Center, School of Aerospace Engineering, Xiamen
University, China. His current research interests include automatic
topology derivation of dc-dc converters and fault-tolerant converters.
Huiqing Wen (M’13-SM’18) received his B.S.
and M.S. degrees in Electrical Engineering from
Zhejiang University, Hangzhou, China, in 2002
and 2006, respectively. In 2009, he received his
Ph.D. in Electrical Engineering from the Chinese
Academy of Sciences, Beijing, China. From 2009
to 2010,he has been an electrical engineer
working with the GE (China) Research and
Development Center Company, Ltd., Shanghai,
China. From 2010 to 2011, he was an engineer at
the China Coal Research Institute, Beijing, China.
From 2011 to 2012, he was a postdoctoral fellow at the Masdar Institute
of Science and Technology, Abu Dhabi, United Arab Emirates. In 2013,
he joined the Electrical and Electronic Engineering Department of Xi’an
Jiaotong-Liverpool University (XJTLU), Suzhou, China. Currently, he is
a senior associate professor at the XJTLU. He has published more than
100 peer reviewed technical papers in leading journals/conferences and
holds over 20 issued/pending patents. His research interests include
renewable energy, electric vehicle, power electronics, Microgrid, and
power semiconductor devices. He is the associate editor of IEEE
ACCESS, International Journal of Photoenergy, and Journal of Power
Electronics.

TII-18-3314
Zheng Wang (S’05–M’09-SM’14) received the
B.Eng. and M.Eng. degrees from Southeast
University, Nanjing, China, in 2000 and 2003,
respectively, and the Ph.D. degree from The
University of Hong Kong, Hong Kong, in 2008, all
in electrical engineering.
From 2008 to 2009, he was a Postdoctoral
Fellow in Ryerson University, Toronto, ON,
Canada. He is currently a full Professor in the
School of Electrical Engineering, Southeast
University, China. His research interests include electric drives, power
electronics, and distributed generation. He has authored or coauthored
over 80 internationally refereed papers and four books in these areas.
Prof. Wang received several academic awards including IEEE PES
Chapter Outstanding Engineer Award, Best Paper Award of
International Conference on Electrical Machines and Systems (ICMES),
Best Session Paper Award of IEEE Annual Meeting of Industrial
Electronics (IECON), and Nanjing Outstanding Paper Award of Natural
Science.
Kai Ni (S’17) was born in Jiangsu, China. He
received the B.Eng. (Hons) degrees in Electrical
Engineering and Automation from Xi’an Jiaotong
Liverpool University, Suzhou, China, and
Electrical Engineering from the University of
Liverpool, Liverpool, UK, in 2016. He is currently
pursuing the Ph.D. degree at the University of
Liverpool. His research interests include operation
and control of doubly-fed induction machines,
power electronic converters, and power systems.