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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
Flexible Voltage Control Strategy Considering
Distributed Energy Storages for DC
Distribution Network
Yong Li, Senior Member, IEEE, Li He, Student Member, IEEE, Fang Liu, Member, IEEE, Canbing Li,
Senior Member, IEEE, Yijia Cao, Senior Member, IEEE, Mohammad Shahidehpour, Fellow, IEEE
 the AC/DC conversion stages [4], [5]; 3) larger power trans-
Abstract—In this paper, a flexible voltage control strategy,
which takes good use of the distributed energy storage (DES) units,
is proposed to enhance the voltage stability and robustness of DC
distribution network. The characteristics of AC/DC interface in
network is analyzed, and the virtual inertia and capacitance are
given to demonstrate the interactive influence of the AC and the
DC systems. The control strategy for DES which is located at the
AC microgrid or at the network terminal bus is designed based on
the interactive characteristics, enabling the DES to respond to
both voltage variation of DC network and frequency change of
utility AC grid. A cascading droop control strategy is proposed for
DES in DC microgrid to relieve the pressure of voltage deteriora-
tion of DC network buses which connect the DC microgrid. The
proposed comprehensive flexible control strategy for DESs at
different interfaces features independence of communication as
well as enhancement of system robustness, and reduces the impact
of DC distribution network on utility AC grid. The performance
of the proposed control strategy is validated under different op-
erating conditions of the DC distribution network.
Index Terms—DC distribution network, voltage control, dis-
tributed energy storage (DES), AC/DC interface, droop control.
I. INTRODUCTION
I N recent years, the development of DC technology has en-
countered a boom in both theoretical and practical regards,
especially in the fields of high voltage direct-current (HVDC)
transmission system and low voltage microgrid [1]-[3]. Com-
pared with the conventional AC system, power systems
adopting DC technology features that: 1) enabling more pene-
tration of the renewable sources and energy storage (ES) ele-
ments, e.g., photovoltaic (PV) systems, batteries, super capac-
itors; 2) improving power efficiency significantly by reducing
This work was supported in part by the national Natural Science Foundation
of China (NSFC) under Grant 61233008, 61673398 and 51520105011, in part
by the Special Project of International Scientific and Technological Coopera-
tion of China under Grant 2015DFR70850, and in part by the S&T Special
Project of Hunan Province of China under Grant 2015GK1002. (Corresponding
authors: Fang Liu and Yijia Cao)
Y. Li, L. He, C. Li and Y. Cao are with the College of Electrical and In-
formation Engineering, Hunan University, Changsha 410082, China (e-mail:
yongli@hnu.edu.cn, helifamily@hnu.edu.cn, lcb@hnu.edu.cn, yjcao@hnu.edu.
cn)
F. Liu is with the School of Information Science and Engineering, Central
South University, Changsha 410083, China (e-mail: csuliufang@csu.edu.cn).
M. Shahidehpour is with the Galvin Center for Electricity Innovation, Illi-
nois Institute of Technology, Chicago, IL 60616 USA. (e-mail: ms@iit.edu).
1949-3053 (c) 2017 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.
1
mission capacity available and problems such as synchroniza-
tion stability, reactive power control and harmonics can be
avoided [6], [7]. With the increasing penetration of renewable
resources and microgrids, the conventional AC distribution
network is facing big challenges in plug-and-play performance
and operating stability, where DC systems have more ad-
vantages. Therefore, given the demand of power system oper-
ation and success of DC technology in some specific applica-
tions, e.g., large-scale data centers, shipboard systems [8], [9],
medium voltage DC (MVDC) distribution network is getting
more attraction in the future smart grid architecture.
Since no concerning factors such as reactive power and
phase synchronization and the like, the DC voltage plays a key
role in the system operation stability. Nowadays, general
voltage control strategies can be mainly classified into two
categories: master-slave control and voltage droop control. In
master-slave control strategy, one voltage source converter
(VSC) is assigned to be slack terminal, which means it is sup-
posed to track the variation of DC voltage and keep the voltage
at the reference value. This control strategy can achieve high
operation accuracy because all the power source converters are
regulated according to the state of the system [10], [11].
However, the master-slave control strategy is tightly dependent
on fast-speed and high-bandwidth communication. Thus, the
redundancy design is required in this control structure. Besides,
this strategy is not friendly to the access of new generation
sources, for the control frame should be refreshed accordingly.
Droop control strategy is free of communication and uses the
voltage signal to regulate the output power of controlled con-
verters [12]-[16]. In [13], a coordinated droop control strategy
is proposed for multi-terminal DC (MTDC) system, which
enables the proportional power dispatch among grid side
HVDC stations. To address the voltage mismatch, an adaptive
droop control method is investigated [14], which is able to
minimize the voltage drop and load current sharing difference
by introducing a figure of merit index. A hierarchical control
strategy is proposed in [15] and [16], which assigns different
control aims into several levels to regulate more involving
elements and mitigate the influence of droop control. However,
it is still difficult to calculate the droop coefficient on the
premise of keeping system stability and avoiding power flow
disorder. The existing master-slave or voltage droop control is
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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
effective for the voltage regulation of the buses near the ter-
minal. However, with the scale expansion of the network, the
voltage of nodes far from the terminal buses will be regulated
more difficultly. Besides, a flexible control strategy is required
to reduce the impact of low-inertia DC system imposed on AC
power grid.
To compensate the mismatch of the distributed sources
generation and the local loads consumption ES technique is
generally utilized [17], [18]. The ESs installed in DC distribu-
tion network are to be distributed, located in the local mi-
crogrids or attached to the important terminal buses of the
distribution network, with the favorable characteristics, i.e.,
easy to be installed, fast response rate [19], [20]. The existing
control strategies for DESs are focused on microgrids, and to
put the ES unit on the role of back-up source or auxiliary ser-
vice. In [21], the ES unit is activated only in the state when the
microgrid is isolated from the utility grid or there is a limitation
for utility grid output, which makes the ES unit of high vacancy
rate. In [22], a coordinated control strategy concerning the net
power allocation with different frequency band is investigated,
which assigns the high frequency power variation to ESs to
track, and is highly dependent on the precise load prediction.
Besides, the aforementioned studies on the ES unit are from the
view of islanding microgrid operation. There is great potential
to involve the ES units in the active power control for
large-scale distribution system such as the DC distribution
network.
In order to realize better power dispatching and management
in smart grid, with the development of smart appliances for load
side, the concept of demand side response (DSR) has been
proposed [23-25]. A DSR for DC distribution network is real-
ized by changing the participating power electronics loads to
maximize the efficiency of renewable resources in [23].
However, the function of ES units to operate as flexible sources
are not fully utilized. In [24], the economic dispatch of active
power for DC microgrid is proposed based on genetic algorithm
(GA), which can minimize the operating cost and is highly
dependent on communication network and the precise load
prediction. To acquire the global optimization of DSR, [25]
gives a new problem formulation of DSR from the system level,
which improves the modelling and search efficacy for hybrid
AC/DC systems. However, the aforementioned DSR methods
Fig. 1. Dual terminal structure of DC distribution network
1949-3053 (c) 2017 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.
2
are mainly considered from the point of the optimal power flow
or economic power dispatching, which features long time in-
terval, rather than improving the short-term dynamic power
quality. Besides, the reported DER methods require a
high-performance central controller and reliable communica-
tion network.
In this paper, a flexible voltage control strategy considering
DESs is proposed for DC distribution network, which is able to
eliminate the node voltage variation during the change of sys-
tem net power, and to improve the inertia response ability of
DC network on AC utility grid. The main contributions of this
paper are summarized as follows:
1) A linearized coupling model between DC voltage and AC
frequency is established, which reveals the interactive influ-
ence of the AC and DC systems. The model is based on the
virtual inertia and virtual capacitance, thus it has an under-
standable physical meaning.
2) A new comprehensive control strategy for DESs is pro-
posed, of which the detailed design varies flexibly according to
the places where the DESs are located, thus the dynamic power
quality can be significantly improved.
3) Compared with existing global optimization methods, the
proposed control strategy is free of communication and feasible
to be implemented into practice, since the strategy is based on
the interactive characteristics between the AC and DC variables
and developed from the common control structure of the con-
verters.
II. TOPOLOGY AND VOLTAGE CONTROL STRATEGY OF DC
DISTRIBUTION NETWORK
A. Topology of DC Distribution Network
Similar to the AC distribution network, the DC network can
be classified into three typical types: 1) the radial structure; 2)
the ring structure; and 3) the dual or the multi-terminal structure
[26]. Fig. 1 shows the typical dual terminal DC network studied
in this paper. To acquire good fault ride-through capability, the
substations at the terminal are connected to 4 kV DC network
via the isolation transformers and the voltage source converter
(VSC), which feature the electric isolation capability and work
together to step down the voltage and transfer AC power into
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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
3

DC format. The following three different elements are inter- Pnet   Pload  ( PDGs   PESs ) (1)
faced into the network via DC cables. where Pload, PDG and PES are the load power, DG power and ES
1) AC/DC microgrid. This type of element usually consists power, respectively.
of distributed generators (DGs), ESs and local loads, the power The control strategy for ES unit of microgrid can be classi-
of which is varied periodically in accordance with the change of fied into two modes. In Mode I, ES unit will not be activated
natural environment factors such as wind speed or photovoltaic during connection, all the net power demands will be satisfied
irradiation. When there is a shortage in power demand, the by absorbing energy from distribution grid, and ES units should
microgrid adjusts the power amount absorbed from the DC take part in power adjustment only when the microgrid is iso-
network. Sometimes, the microgrid can be used to adjust the lated from grid, or in the situation that power flow is beyond the
voltage of distribution network. capacity limitation, denoted as PNIC, of ICs [21]. This mode can
2) AC/DC loads. The aggregative loads feature unidirec- be described by
tional power flow and can hardly be taken into consideration
 Pdcref   Pload   PDGs and  PESs  0, when 0  Pnet  PICN

for voltage control. Except for some emergency condition, the  ref (2)
 Pdc  PIC and  PESs  Pnet  PIC , when Pnet  PICN
N N
loads can be shed in a passive way to release the burden of 
network. where Pref
dc is the reference power flowing from the DC distri-

3) Independent ES unit. It can be installed at any node, and bution network to the microgrid.
offer an ancillary or backup support for DC voltage. The signal In Mode II, it features that ES unit is assigned to track the
of node voltage is collected as the input for ES unit controller to net power variation of microgrid [22]. The ES unit is utilized
compensate the voltage variation. to absorb high-frequency power variation, and the rest of
power vacancy is compensated by the DC distribution net-
B. Conventional Control Strategy for Network Bus Voltage
work. In some cases, the ES unit is responsible for all of the
According to the classification of network elements men- power vacancy which is not beyond the operational limits of
tioned above, the nodes connected with different types of ele- ES converter, and DC network will only take part in the
ments show different operating characteristics. Fig. 1 shows the power adjustment when the power burden is too much for ES
control strategies of different types of nodes in the DC distri- to handle. This mode can be expressed by
bution network. More specifically, in this section, the voltage  1
 ESs
control of the nodes with different elements will be investi-  P  Pnet Glpf ( s )  Pnet
1 s (3)
gated. 
 P ref  P (1  G ( s ))  P  s
1) Terminal nodes. At the terminal of DC network, the  1 s

dc net lpf net

AC/DC converters are utilized to offer an access to AC grid for or

the network. These converters work under one of three control
 PESs  Pnet
strategies, namely, the constant voltage control, the droop 
control (V-P) or the constant power control, as shown in in Fig.  ref 0, if Pnet   PES , max (4)
 Pdc  
 Pnet   PES , max , if Pnet   PES , max
2. No matter which topology the DC network adopts, at least

one of the terminal converters should adopt the constant voltage
control to ensure there is a slack terminal in the system [27]. where PES,max is the maximum output power of a ES unit in
microgrid; and τ is the time constant of the first-order low-pass
Constant voltage control Voltage droop control Constant power control
filter .
Vref I dc Vref I dc Pref I dc
PI 1/k PI For Mode I, ES units are of high vacancy rate. For Mode II,
Vdc Vdc Pdc Pdc the capability of power support of distribution network is not
S=1 S=2 S=3 fully utilized, which means the ES units are of high pressure on
I dcref Current
ref
Vabc power adjustment. Besides, the conventional control strategy
PWM
controller for DC distribution network mainly focuses on the voltage
VSC quality by tuning the terminal VSCs, which means the DC
Fig. 2 Different voltage control strategies for terminal converter of DC distri- network has no response to the frequency variation of AC
bution system network. Therefore, the conventional strategy is not friendly to
2) The nodes connected with aggregating loads. These nodes the utility grid and indirectly deteriorates the frequency stability,
work in a constant power consuming mode. In some emergency and it would be become harder and more complex to keep the
cases, some loads may increase the power demand within a global voltage quality when the scale of DC distribution net-
narrow range. work enlarges and more nodes are included.
3) The nodes connected with microgrids. The microgrids To overcome these disadvantages, based on the principle of
connecting to DC distribution network can output power if the droop characteristics, the control strategies for DESs near dif-
distributed generators have more power that the local loads ferent interfaces are designed and will be discussed in section
cannot be consumed, and the interface converters (ICs) are III.
controlled in the similar way as the terminal VSC. When there
is a lack of power demand in microgrid, its net power Pnet is III. FLEXIBLE VOLTAGE CONTROL WITH DESS
defined as In this paper, the comprehensive demands of the operating

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
4

characteristics and control aims for different interfaces are 2 H vir df CvirVdc , k dVdc , k
taken into account. As shown in Table I, the power flowing  (7)
f 0 dt S IC dt
from microgrids or AC utility grid to the DC network is con- AC/DC
sidered in the condition of charge, and thus the inverse power Pin Pout Coupling PE
flow is in the discharge state. The inner traits describe the var- +
PC
iation of the electrical characteristics of elements interfaced to Pf
Vdc f ,
the DC network, such as the frequency of AC grid and the PM
voltage of DC grid. The control objectives of this paper are
described in this Table, where the absolute frequency deviation -
|Δf | is limited within 1%, and the DC voltage variation is within Fig. 3 Equivalent power relationship for AC/DC interface
5%. ΔVdc and ΔVbus represent the voltage variation of DC mi- In order to reveal the interactive power supporting capability
crogrid and DC distribution network, respectively. of both AC and DC sides, (7) can be integrated as
It should be noted that when some severe power events occur
f 2 H vir Vdc ,k CvirVdc , k
and the power quality cannot be guaranteed, which means the
above delta values are beyond their limits, the common emer-
 f0 f0
df  
V ref S IC
dVdc , k (8)

gency measures like load shedding or microgrid disconnection 2 H vir C

will be activated to keep the stability of the DC distribution ( f  f 0 )  vir (Vdc2 , k  Vref2 )
f0 2S IC
system [28]. (9)
TABLE I Vdc , k  M ( f  f 0 )  V 2
ref
(10)
CHARACTERISTICS OF DIFFERENT INTERFACE OPERATION AND CONTROL 4 H vir S IC
OBJECTIVES M
Cvir f 0
Type of inter- Control (11)
Power flow Inner traits Node traits
face objectives where f0 and f represent the initial value of the AC frequency
Frequency Voltage and the stable value of frequency after power disturbance in
AC/DC Charge
increase decrease |Δf |<1%
(AC microgrids AC/DC interface, respectively; M is defined as the correlation
Frequency Voltage &|ΔVbus|<5%
/ utility grid) Discharge coefficient of the DC voltage and the AC frequency.
decrease increase
Voltage Voltage Equation (10) can be rewritten in another form, in order to
DC/DC Charge
increase decrease |ΔVdc|<5% demonstrate the relationship of DC voltage change and AC
(DC mi-
Voltage Voltage &|ΔVbus|<5% frequency variation, i.e.,
crogrids) Discharge
decrease increase 4 H vir ( f  f 0 ) S IC 1
Vdc , k  Vref 
A. Interactive power balance analysis for AC&DC grid f 0 Cvir Vdc , k  Vref
(12)
For the AC system, the power balance can be described by Substituting f-f0=Δf and Vdc,k-Vref=ΔV into (12) yields
the swing equation derived from the dynamic characteristics of 4 H vir S IC f
V 2  2Vref V 
synchronous machine, i.e., f 0 Cvir
(13)
2 H SG df
 PM  PE  Pf (5) 4 H vir S IC f
f 0 dt V  Vref   Vref2 (14)
f 0Cvir
where HSG is the inertial constant of synchronous generator; PM
and PE are the input mechanical power and the output elec- As can be seen from (14), the variation of DC voltage is
tromagnetic power, respectively. nonlinear with the frequency change, and these two variables
For the DC grid, the power dynamic characteristics are de- are linked through the parameters Hvir and Cvir. Fig. 4 gives the
termined by the DC bus voltage. The power balance of DC side visualization of (Δf-ΔV) characteristics based on the parameters
of AC/DC interface can be expressed by of DC network tested in section IV. The figure shows that with
smaller Cvir, the larger ΔV is required to provide the same-level
CdcVdc , k dVdc , k
=Pin  Pout  PC (6) inertial constant, which means the inertia characteristics can be
S IC dt provided through DC system by releasing more active power
where Cdc is the capacitance of the DC side capacitor; SIC is the (corresponding to more obvious voltage drop).
rated power capability of IC; Vdc,k is the voltage of the kth bus in
the DC network; and Pin and Pout are the input power of DC side
and the output power of AC side, respectively.
When the power is transferred from DC grid to AC grid, the
IC can be treated as a virtual synchronous machine seen from
AC side, providing the frequency support [29]; similarly, when
the power is absorbed from AC grid to DC grid, the IC can be
treated as a virtual capacitor, providing the voltage support.
Such interaction of power balance is illustrated in Fig. 3, and
combining (5) and (6) yields Fig. 4 Relationship among virtual inertial constant Hvir, virtual capacitance Cvir,
frequency and voltage variation (Δf, ΔV)

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
5

B. Voltage control strategy for AC/DC interface

When there is lack of power in AC microgrid, the power
vacancy will be compensated by the DC network. When the
droop control method is adopted, it can be described as
1
Pref   ( f  f 0 )  P0 (15)
kac
where P0 and Pref represent the initial output power and the
reference output power transferred by IC to AC microgrid,
respectively; and kac is the droop coefficient of AC microgrid.
Linearizing (15) by using Taylor series, one can obtain Fig. 5 Relationship among ES output power, virtual inertia and voltage varia-
tion (DC capacitance is 0.45mF, the rated power of ES unit is 3 MW, and P0ac,ES
M M2
V  f  3 f 2  o(f 2 ) (16) is set equal to 0 MW)
2Vref 8Vref
where o(.) represents the higher order infinitesimal terms of Δf.
Considering that, in practice, the frequency variation is rela-
tively small (restricted within ±0.1p.u.), it is reasonable to keep
only the steady-state dc term and the first-order term of (16),
which leads to
M
Vdc , k  Vref  ( f  f0 ) (17)
2Vref
Equation (17) gives the linearized coupling model between
DC voltage and AC frequency, which reveals the interactive
influence of the AC and DC systems. When the DC network is
Fig. 6 Relationship between kac and Pac,ES (the rated power of ES unit is 3 MW
under the extreme condition such as isolation and power limi- and ΔV is equal to -0.3kV)
tation, the net power of AC microgrid should be compensated Based on the general control structure of VSC [30], the
in part or all by its local ES units. Thus, substituting (15) into control diagram of DC/AC converter for the ES unit in AC
(17) yields microgrid is developed as shown in Fig. 7. The hysteresis con-
2Vref troller is adopted to avoid the frequent activation of the ES unit
Pacref, ES   (Vdc , k  Vref )  Pac0 , ES
Mkac (18) when the DC voltage changes near the boundaries of dead band,
 kb (Vref  Vdc , k )  P 0
ac , ES which can protect the ES unit and prevent the voltage oscilla-
where P 0
and P
ac,ES
ref
represent the initial and the reference
ac,ES
tion due to the power compensation. Note that the droop control
output power of AC microgrid ES unit, respectively. The con- method described in (18) will not change the trend of voltage
stant kb is defined as the equivalent droop coefficient of ES unit. variation of the DC network, thus no power flow disorder will
When there is a voltage variation of DC bus, resulted from be caused.
the power compensation for AC grid, the ES unit of AC mi-
crogrid will be activated to output power to the local microgird, Outer controller

and to support the DC voltage. As can be known from (11) and Vref Vdc,k
Low pass filter
Common AC bus
the definition of kb, the output power of ES unit is determined ΔV Other elements kthDC bus
by virtual inertial constant Hvir and the DC voltage variation ΔV. iES
VSC
iL

The capacitance Cdc and the droop coefficient kac are set as the hysteresis
controller
ES
unit
VES
Reactor
Vdc,k

fixed values, so they have no influence on the output power. Filter

Equation(18) ICAC/DC
Pac0 , ES
Fig. 5 shows the relationship curve among the ES output Pacref, ES PWM DC distribution
power, the virtual inertia and the voltage variation. With the 1/VES network
ref
i ac , ES
Modulating signal
idref
increase of the voltage deviation, the ES unit outputs or absorbs iES
- PI PI
in d_q frame
AC/DC
dq interface
more power. However, the rise of Hvir imposes a converse Qref
i ref
q
abc
Q - PI PI
effect on the output power of ES unit. In a certain extreme Inner controller

condition, e.g., ΔV=-0.1p.u, the maximum deference between Q: reactive power AC microgrid
idref : reference current of d-axis
the ES output power references can be as large as 0.3 p.u. Qref: reference reactive power
iqref : reference current of q-axis
iES: output current of ES unit
Therefore, to prevent ES unit from over charging or over dis-
Fig. 7 Controller configuration for the ES unit in AC microgrid
charging, Hvir should be adjusted according to the characteristic
curve shown in Fig. 5. Further, as can be seen in Fig. 6, when By adopting the aforementioned control strategy, the ES unit
ΔV is set as a fixed value (e.g., -0.3kV), the increase of kac will in AC microgrid is able to respond to the voltage variation of
significantly lower the ES power, especially for larger Hvir. It the bus where the microgrid accesses. The output power of ES
means that the determination of Hvir can vary in a wider range if unit is proportional to the voltage deviation. Since the voltage
the AC microgrid is able to bear larger frequency variation deviation is traced and compensated, the stability of DC net-
(corresponding to larger kac). work can be enhanced.

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
6

In order to take into account the influence of the DC network grid. Fig. 8 shows the comprehensive control diagram for the
side voltage variation on the AC frequency, the voltage devia- terminal AC/DC interface.
tion component ΔVdc_f, which is denoted by the frequency DC distribution
Outer controller
variation, is defined as follows: network Terminal DC bus
f - f0
M 2 H vir S IC
Vdc _ f  ( f  f0 )  ( f  f0 )
Other elements
(19) Δf
2Vref Cvir f 0Vref iES
DC/DC
hysteresis
In (19), the inertial constant Hvir included in the V-f correla- controller
ES VES
Vdc,k
tion coefficient M should be determined according to the Equation(19)
unit

practical characteristics of the AC grid, also denoted by HSG S ΔVdc_f 0

PES
in this situation. Therefore, Hvir helps to transfer Δf of utility
1 2 _f
iES PWM
PI 1/ku

grid to the voltage deviation, denoted by ΔVdc_f, in DC network,

ref ref
PES _f i ES -
1/VES PI
thus the DC network can indirectly contribute to the inertia Inner controller PI

response by responding to ΔVdc_f using the independent termi- PLL PI

- i ref PI
ΔQ
Outer
iq q controller Vref
abc
nal ES units. Besides, the value of Cvir is provided by the dq
id - - 1
system operators, which inversely affects the value of ΔVdc_f.
ref
Inner
i d P ref
dc PI S
PI PI
controller 1/ku -
Combining (9) and (19) yields PWM
P0
dc
2 Vdc,k

f
4 H SG S IC Reactor
Vdc,k
f0 (20)
Cvir  Utility grid Trans
 Vdc _ f  Filter
Terminal
Vref2 (  1) 2  1
VSC
 Vref 
It can be seen that, the larger inertial constant HSG leads to Fig. 8 Comprehensive control diagram for the terminal AC/DC interface
larger virtual capacitance Cvir, and corresponds to lower ΔVdc_f, C. Voltage control strategy for DC/DC interface
which is in accordance with the physical characteristics of
The bi-directional DC/DC interface is located between the
capacitor. Besides, a large value of HSG ensures smaller Δf, and
DC distribution network and the DC microgrid, where the
requires less power support for frequency stability. In practice,
high-power bi-directional DC/DC converter is installed as IC.
Cvir is calculated by the allowable maximum frequency devia-
For simplifying the analysis, it is assumed that voltage droop
tion Δfmax and the maximum voltage variation ΔVmax dc .
control method is adopted to regulate the DC microgrid voltage,
When the voltage fluctuation of terminal bus appears, the
which can be described as
utility grid will transfer power through VSC to the DC network
1
according to the predefined voltage droop or the constant Pdcref, IC   (VMG  VMG
ref
)  Pdc0 , IC (23)
voltage control strategy. Meanwhile, the independent ES unit, kdc
used to regulate voltage in emergency condition, is activated to where P0dc,IC and Pref
dc,IC represent the initial output power and the

track the voltage deviation component ΔVdc_f. The mathemati- reference output power transferred by IC to the DC microgrid,
cal description of the reference output power of the utility grid respectively; and Vref MG and VMG are the reference and the meas-

and the independent ES unit is shown as follows: ured voltage of DC microgrid, respectively; kdc is the droop
 Pdcref  kvp (Vref  Vdc , k  Vdc _ f )  kvi (Vref  Vdc , k  Vdc _ f )  Pdc0 coefficient, which is preset by the DC microgrid operator.
 
 (21) Unlike the conventional control strategy, the ES is assigned
 ref  M   M 
 PES _ f  kvp  ( f 0  f )   kvi   ( f 0  f )  PES
0
_ f
to track the voltage of DC network side bus where the DC

  ref
2V   ref
2 V  microgrid is located, rather than the voltage of DC microgrid
 ref 1 itself, and thus a cascading power compensation method is
 Pdc  k (Vref  Vdc , k  Vdc _ f )  Pdc
0

 u (22) developed and described as

 1
 PESref_ f   M ( f  f 0 )  PES0 _ f

 2kuVref
where kvp and kvi represent the proportional and the integral
coefficients of the outer PI controller, respectively; ku is the
droop coefficient for the terminal VSC.
Equation (21) is obtained with the constant voltage control
strategy while (22) is derived with droop control. The reference
output power of the independent ES unit is determined by
ΔVdc_f and added as the supplementary control input of the
Pdcref, ES   (Vdc , k  Vref )  Pdc0 , ES (24)
kES
where kES is the droop coefficient for the ES unit of DC mi-
crogrid.
Given the worst scenario in which the largest voltage sag of
DC microgrid appears, and the bus voltage of DC network
approaches near the acceptable lowest boundary, the ES should
guarantee the stability of the DC bus and the microgrid.
Therefore, the allowable largest voltage deviation ΔVmaxMG of the

terminal VSC signal, serving as a modified term to regulate the DC microgrid, along with the allowable largest voltage devia-
power absorbed from the utility grid. By using this control tion ΔVmax
dc,k of the DC distribution network side bus, are used to

method, the DC network is able to respond to the frequency calculate kES expressed by
variation of the utility grid via the ES unit, without influencing Vdcmax (25)
k ES  ,k

the effect of voltage control strategy for the DC network. Thus, 1

VMG
max
 Pdc0 , IC  Pdc0 , ES
the DC network can perform as a friendly load to the AC utility kdc

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
7

IV. CASE STUDIES of bus voltage, the output power of ES unit is reduced to 0.1
In order to validate the proposed voltage control strategy, a MW, as shown in Fig. 9(d).
comprehensive simulation model of DC network based on Fig. 2) Increase of Net Power of DC Microgrid: When the net
1 is developed. The independent ES unit is installed at the power increment of 0.3 MW is started at t=4s, the voltage of
terminal bus denoted #1 of the DC network, and the microgrids DC microgrid drops from 200 V to 197.5 V, and the active
are also equipped with ES units. To acquire a good balance power is compensated via the IC by absorbing more power
between energy density and fast power response, the Li-on from the capacitor at the DC side, resulting in the sudden de-
batteries are chosen as the ES units in the simulation cases [16], crease of bus voltage. The DESs in the AC and DC microgrids
with the power ramp rate of 1.5 MW/s, and the capacities are are controlled by the proposed strategy described in (18) and
listed in Table II, respectively. Aggregating AC/DC loads are (24) to regulate the voltage of respective DC bus. It should be
substituted by a constant power load, since they play no sig- noted that, as the frequency of utility grid connected to VSC1
nificant part in the following simulation. The line resistances of drops under 49.8 Hz, the independent ES unit is activated to
each DC cables are R1=0.1Ω, R2=0.25Ω, R3=0.05Ω, respond to the frequency change, as shown in Fig. 11(d).
R4=R5=0.1Ω. Other system parameters are given in Table II. 3) Decrease of Net Power of AC Microgrid: When the net
power reduction of 0.6 MW in AC microgrid is started at t=5 s,
TABLE II
the frequency of AC microgrid returns to near 50 Hz. Due to the
PARAMETERS OF THE STUDIED DC DISTRIBUTION NETWORK
Sub-systems Items Values
decrease of power demand, the voltage of every DC bus is
Rated power/MVA 5
increased, and those ES units which correspond to the voltage
Nominal frequency/Hz 50
variation reduce the output power accordingly, as shown in Fig.
Utility grid
(synchronous generator) Droop coefficient/(Hz/MW) 0.39
9 (d) and Fig. 10 (d)
Inertial constant/s 0.6
4) Disconnection of Non-Slack Terminal: The external utility
Nominal voltage/kV 4 grid at the terminal VSC2 is disconnected from DC network at
DC
Allowable bus voltage variation(p.u.) ±0.02 t=7 s, which means that the distribution network is only sup-
distribution network ported by the terminal VSC1. As shown in Fig. 9(a) and Fig.
Rated capacity of ES unit/(kWh) 20
Droop coefficient/(Hz/MW) 0.3 10(a), the DC voltage is down to the lower limit 3.6 kV. The
Capacitor of DC side/mF 0.46 frequency of utility grid connected to VSC1 drops to nearly
Rated power of the IC/MVA 3 49.5 Hz, as shown in Fig. 11(b). If some unexpected disturb-
AC microgrid Voltage dead band of ES(p.u.) ±0.0125 ances are brought to utility grid, the DC network may be iso-
Virtual inertial constant/s 6.5×10-3 lated to guarantee the security of the external power grid. In
Nominal voltage/V 200 comparison, with the proposed DESs control strategy, the
Rated capacity of ES unit/(kWh) 10 voltage only drops to approximately 3.9kV, and the frequency
Droop coefficient/(V/MW) 25 margin is increased effectively for the utility grid.
Voltage dead band of ES(p.u.) ±0.0125 It can be seen from Figs. 9(e) and 10(e) that, with the pro-
DC microgrid
Droop coefficient of ES/(V/MW) 60 posed strategy, the DESs are supposed to output fewer active
Rated capacity of ES unit/(kWh) 10 power than that of Mode II, corresponding to smaller change of
state of charge (SoC), which not only contributes to the dy-
A. Case 1: Dynamic Performance on Power Variation namic power quality, but also prolongs the ES service life.
This case is used to test the proposed control strategy on the
power variation appearing in DC network. The simulation re-
sults are shown in Figs. 9-11, and the DC voltage changes of
terminal VSCs are given in Table III. It should be noted that,
with DESs scenario means the proposed strategy is adopted,
whereas without DESs corresponds to the strategy of Mode I, as
illustrated in Section II.B, in which the ES unit will not be ac-
tivated during the connection of microgrids. Besides, Full-DESs
support represents Mode II, in which the whole variation of net
power is compensated by ES units based on the precise load
prediction, thus no significant variation of DC voltage will be
caused.
1) Increase of Net Power of AC Microgrid: When the net
power increment of 0.6 MW is started at t=2 s, which leads to
the decrease of frequency approaching 49.8 Hz, the IC transfers
more power from the DC network to retain the stability of the
AC microgrid. Instant DC bus voltage decrease of 130 V
(0.0325 p.u.) is detected, as shown in Fig.9 (a), and the ES unit
in AC microgrid outputs 0.25 MW active power to compensate
the voltage variation, according to (18). With the stabilization

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
8

Fig.9 Simulation results of the AC microgrid for case 1, (a) voltage of DC bus
#3; (b) frequency variation; (c) change of net power; (d) output power of the ES
unit; (e) SoC of ES unit

Fig.11 Simulation results of DC bus #1 for case 1, (a) voltage of DC bus; (b)
frequency variation of the utility grid; (c) active power of VSC1; (d) output
power of the independent ES unit

TABLE III
CHANGE OF THE TERMINAL DC VOLTAGE (CASE 1)
Terminal Condition 0-2 s 2-4 s 4-5 s 5-7 s 7-9 s
With DESs 4 3.93 3.96 4.10 3.88
VSC1 (kV)
Without DESs 4 3.86 3.92 4.13 3.76
With DESs 3.96 3.90 3.90 4.01 3.76
VSC2 (kV)
Without DESs 3.96 3.81 3.84 4.10 3.63

2) Cable Fault: A disconnection on the DC cable between

the nodes of the AC and the DC microgrids occurs at t=4 s. The
DC network is separated into two individual parts. When de-
Fig.10 Simulation results of the DC microgrid for case 1, (a) voltage of DC bus tecting the serious fault, both terminal VSCs switch to constant
#4; (b) voltage variation; (c) change of net power; (d) output power of the ES voltage control mode. As shown in Fig.12 (a) and Fig.13 (a),
unit; (e) SoC of ES unit the voltage variation is smoothed with the participation of
DESs. Besides, the power demand by the DC network is au-
B. Case 2: Dynamic Performance on VSC Mode Change and tomatically adjusted according to the frequency change of the
Cable Fault utility grid, as shown in Fig. 12(b). Therefore, the stability of
This case is used to test the proposed control strategy on the the whole system involving DC network and the utility grid can
change of the operating mode of the terminal VSC, and the be enhanced.
serious outage fault happens in the network. The simulation
results are shown in Figs. 12-14, and the DC voltages of ter-
minal VSCs are given in Table IV.
1) VSC Mode Change: When the net power increment of 1
MW is activated at t=2 s, VSC1 outputs the fixed active power
of 2.3 MW and lacks the ability of voltage regulation. When
there is no DES response, VSC2 switches to the droop control
mode to respond to the voltage sag, and inject 0.15 MW power
into the DC network, as shown in Fig. 13(b). In comparison,
with the DESs, the operating mode of VSC2 is not changed,
since the net power vacancy is compensated by the DESs in
microgrids. Besides, the voltages of terminal buses are beyond
Fig.12 Simulation results of DC bus #1 for case 2, (a) voltage of DC bus; (b)
3.8 kV during the power change, as shown in Fig. 12(a) and Fig. active power of VSC1
13(a).

1949-3053 (c) 2017 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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
Fig.13 Simulation results of DC bus #2 for case 2, (a) voltage of DC bus; (b)
active power of VSC2
Fig.14 Output power of the ES units in microgrids for case 2
TABLE IV
CHANGE OF THE TERMINAL DC VOLTAGE (CASE 2)
Terminal Condition 0-2 s 2-4 s 4-6 s
With DESs 4 3.89 4.02
VSC1 (kV)
Without DESs 4 3.75 4.07
With DESs 3.96 3.85 3.76
VSC2 (kV)
Without DESs 3.96 3.71 3.64
V. CONCLUSION
In this paper, a flexible voltage control strategy, which fully
considers the regulation ability of the DES units in DC distri-
bution network, is proposed. The proposed strategy sufficiently
utilizes the features of different interfaces in the DC network,
and shows good performance in diminishing the voltage varia-
tion of DC buses. The relationship between the AC frequency
and the DC voltage is identified via the interactive virtual in-
ertia and capacitance, which is used to design the voltage con-
trol strategy for AC/DC interface. The strategy enables the DES
units in AC microgrid to relieve the voltage variation of DC
network, and improves the system stability by the help of in-
dependent ES unit to respond to the frequency of the AC utility
grid. Besides, a cascading droop control method is involved to
provide the power support for voltage stability of the interface
between the DC network and the DC microgrid. The perfor-
mance of the proposed strategy is validated by considering
different scenarios in the case studies. In the future works, the
hybrid ES system will be considered for the more complex
AC/DC distribution networks.
REFERENCES
[1] P. Bresesti, W. L. Kling, R. L. Hendriks, and R. Vailati, “HVDC connec-
tion of offshore wind farms to the transmission system,” IEEE Trans.
Energy Convers., vol. 22, no. 1, pp. 37-43, Mar. 2007.
[2] M. Aragüés-Peñalba, A. Egea-Alvarez, O. Gomis-Bellmunt, and A.
Sumper, “Optimum voltage control for loss minimization in HVDC mul-
ti-terminal transmission systems for large offshore wind farms,” Electric
Power Systems Research, vol. 89, pp. 54–63, Aug. 2012.
[3] H. Kakigano, Y. Miura, and T. Ise, “Low-voltage bipolar-type DC mi-
crogrid for super high quality distribution,” IEEE Trans. Power Electron.,
vol. 25, no. 12, pp. 3066–3075, Dec. 2010.
1949-3053 (c) 2017 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.
9
[4] S. Sanchez and M. Molinas, “Degree of influence of system states transi-
tion on the stability of a DC microgrid,” IEEE Trans. Smart Grid, vol. 5,
no. 5, pp. 2535–2542, Sep. 2014.
[5] P. Shamsi and B. Fahimi, “Dynamic behavior of multiport power elec-
tronic interface under source/load disturbances,” IEEE Trans. Ind. Elec-
tron., vol. 60, no. 10, pp. 4500–4511, Oct. 2013.
[6] R. S. Balog, W. W. Weaver, and P. T. Krein, “The load as an energy asset
in a distributed DC smart Grid architecture,” IEEE Trans. Smart Grid, vol.
3, no. 1, pp. 253–260, Mar. 2012.
[7] D. Salomonsson and A. Sannino, “Low-voltage DC distribution system for
commercial power systems with sensitive electronic loads,” IEEE Trans.
Power Del., vol. 22, pp. 1620–1627, Jul. 2007.
[8] G. AlLee and W. Tschudi, “Edison redux: 380 Vdc brings reliability and
efficiency to sustainable data centers,” IEEE Power Energy Mag., vol. 10,
no. 6, pp. 50–59, Nov./Dec. 2012.
[9] J. C. Ciezki and R. W. Ashton, “Selection and capability issues associated
with a navy shipboard and DC zonal electric distribution,” IEEE Trans.
Power Del., vol. 15, no. 2, pp. 665–669, Apr. 2000.
[10] M. Farhadi and O. Mohammed, “Real-time operation and harmonic
analysis of isolated and non-isolated hybrid DC microgrid,” IEEE Trans.
Ind. Appl., vol. 50, no. 4, pp. 2900–2909, Jul./Aug. 2014.
[11] J. Zhu and C. Booth, “Future multi-terminal HVDC transmission systems
using voltage source converters,” in Proc. Int. Univ. Power Eng. Conf.,
Cardiff, U.K., Sep. 2010, pp. 1–6.
[12] J. Guerrero, L. Hang, and J. Uceda, “Control of distributed uninterruptible
power supply systems,” IEEE Trans. Ind. Electron., vol. 55, no. 8, pp.
2845–2859, Aug. 2008.
[13] L. Xu and L. Yao, “DC voltage control and power dispatch of a mul-
ti-terminal HVDC system for integrating large offshore wind farms,” IET
Renew. Power Gener., vol. 5, no. 3, pp. 223–233, May. 2011.
[14] S. Augustine, M. K. Mishra, and N. Lakshminarasamma, “Adaptive droop
control strategy for load sharing and circulating current minimization in
low-voltage standalone DC microgrid,” IEEE Trans. Sustain. Energy, vol.
6, no. 1, pp. 132–141, Jan. 2015.
[15] J. M. Guerrero, J. C. Vasquez, J. Matas, L. G. de Vicuna, and M. Castilla,
“Hierarchical control of droop-controlled ac and dc microgrids—A gen-
eral approach toward standardization,” IEEE Trans. Ind. Electron., vol. 58,
no. 1, pp. 158–172, Jan. 2011.
[16] JF. Xiao, P. Wang, et al. “Multilevel Energy Management System for
Hybridization of Energy Storages in DC Microgrids,” IEEE Trans. Smart
Grid, vol.7, no.2, pp.847-856, Mar. 2016.
[17] L. Feng, J. Zhang, G. Li, et al. “Cost reduction of a hybrid energy storage
system considering correlation between wind and PV power,” Prot. Con-
trol Mod. Power Syst, vol.1, no.1, pp.1-11, Dec.2016.
[18] Sebastián, R., and R. P. Alzola. “Effective active power control of a high
penetration wind diesel system with a Ni–Cd battery energy storage,”
Renewable Energy, vol. 35, no. 5, pp.952-965, May. 2010.
[19] X. Y. Chen et al., “Integrated SMES technology for modern power system
and future smart grid,” IEEE Trans. Appl. Supercond., vol. 24, no. 5,
pp.1-5, Oct. 2014.
[20] G. Wang, M. Ciobotaru, and V. G. Agelidis, “Power smoothing of large
solar PV plant using hybrid energy storage,” IEEE Trans. Sustain. Energy,
vol. 5, no. 3, pp. 834–842, Jul. 2014.
[21] B. Wang, M. Sechilariu, and F. Locment, “Intelligent DC microgrid with
smart grid communications: Control strategy consideration and design,”
IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 2148–2156, Dec. 2012.
[22] B. Liu, F. Zhuo, Y. Zhu, and H. Yi, “System operation and energy man-
agement of a renewable energy-based DC nano-grid for high penetration
depth application,” IEEE Trans. Smart Grid, vol. 6, no. 3, pp. 1147–1155,
May 2015.
[23] H. Mohsenian-Rad and A. Davoudi, “Towards building an optimal de-
mand response framework for DC distribution networks,” IEEE Trans.
Smart Grid, vol. 5, no. 5, pp. 2626–2634, Sep. 2014.
[24] C. Li, F. D. Bosio, et al. “Economic dispatch for operating cost minimi-
zation under real-time pricing in droop-controller DC microgrid,” IEEE J.
Emerg. Sel. Topics Power Electron., vol. 5, no. 1, pp. 587–595, Mar. 2017.
[25] C. Zhao, SF. Dong, CH. Gu, et al. “New problem formulation for optimal
demand side response in hybrid AC/DC systems,” IEEE Trans. Smart Grid,
DOI: 10.1109/TSG.2016.2628040.
[26] J. J. Justo, F. Mwasilu, J. Lee, and J.-W. Jung, “AC-microgrids versus DC
microgrids with distributed energy resources: A review,” Renew. Sustain.
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/TSG.2017.2734166, IEEE
Transactions on Smart Grid
10

Energy Rev., vol. 24, pp. 387–405, Aug. 2013.
[27] C. Gavriluta, J. I. Candela, C. Citro, J. Rocabert, A. Luna, and P. Rodri-
guez, “Decentralized primary control of MTDC networks with energy
storage and distributed generation,” IEEE Trans. Ind. Appl., vol. 50, no. 6,
pp. 4122–4131, Apr. 2014.
[28] F. Nejabatkhah and Y. W. Li, “Overview of power management strategies
of hybrid ac/dc microgrid,” IEEE Trans. Power Electron., vol. 30, no. 2,
pp. 7072–7089, Dec. 2015.
[29] J. Zhu, C. D. Booth, G. P. Adam, A. J. Roscoe and C. G. Bright,
"Inertia Emulation Control Strategy for VSC-HVDC Transmission
Systems," IEEE Trans. Power Syst., vol. 28, no. 2, pp. 1277-1287,
May 2013.
[30] W. Guo, L. Mu, “Control principles of micro-source inverters used in
microgrid,” Prot. Control Mod. Power Syst, vol.1, no.1, pp.1-7, Dec. 2016.
Yong Li (S’09–M’12–SM’14) was born in Henan,
China, in 1982. He received the B.Sc. and Ph.D.
degrees in 2004 and 2011, respectively, from the
College of Electrical and Information Engineering,
Hunan University, Changsha, China.
Since 2009, he worked as a Research Associate at
the Institute of Energy Systems, Energy Efficiency,
and Energy Economics (ie3), TU Dortmund University,
Dortmund, Germany, where he received the second Ph.
D. degree in June 2012. After then, he was a Research
Fellow with The University of Queensland, Brisbane,
Australia. Since 2014, he is a Full Professor of electrical engineering with
Hunan University. His current research interests include power system stability
analysis and control, ac/dc energy conversion systems and equipment, analysis
and control of power quality, and HVDC and FACTS technologies.
Li He (S’16) was born in Hunan, China, in 1991. He
received the B.Sc. degree in electrical engineering
from Changsha University of Science and Technology
in 2013. He is currently working toward the PH.D.
degree of electrical engineering in the College of
Electrical and Information Engineering, Hunan Uni-
versity, Changsha, China.
His research interests include the control of hybrid
AC/DC power system and the application of energy
storages.
Yijia Cao (M’98-SM’2013) was born in Hunan,
China, in 1969. He graduated from Xi'an Jiaotong
University, Xi'an, China in 1988 and received M.Sc.
degree from Huazhong University of Science and
Technology (HUST), Wuhan, China in 1991 and
Ph.D from HUST in 1994. From September 1994 to
April 2000, he worked as a visiting research fellow,
research fellow at Loughborough University, Liv-
erpool University and University of the West Eng-
land, UK. From 2000 to 2001, he was employed as a
full professor of HUST, and from 2001 to 2008, he
was employed as a full professor of Zhejiang University, China. He was ap-
pointed deputy dean of college of Electrical Engineering, Zhejiang University
in 2005. Currently, he is a full professor and vice president of Hunan University,
Changsha, China.
His research interests are power system stability control and the application
of intelligent systems in power systems.
Mohammad Shahidehpour (F’01) is the Bodine
Chair Professor and Director of Robert W. Galvin
Center for Electricity Innovation at Illinois Institute
of Technology, Chicago, IL, USA. He is a Research
Professor at King Abdulaziz University in Jeddah,
Saudi Arabia, North China Electric Power University
in Beijing, and Sharif University in Tehran.
Dr. Shahidehpour is the recipient of the Honorary
Doctorate in 2009 from the Polytechnic University
of Bucharest in Romania. He is an IEEE Distin-
guished Lecturer, Chair of the 2012 IEEE Innovative Smart Grid Technologies
Conference, Chair of the 2012 Great Lakes Symposium on Smart Grid and the
New Energy Economy. He is the recipient of the 2012 IEEE PES Outstanding
Power Engineering Educator Award.

Fang Liu (M’12) was born in Jiangxi, China, in 1982.

She received the B.S. degree from the College of
Electrical and Information Engineering, Zhengzhou
University of Light Industry, Zhengzhou, China in
2005 and received the M.S. degree from the School of
Information Science and Engineering, Central South
University, Changsha, China in 2007. In 2008, she
became a Ph.D. student in Waseda University, Japan
and received the Ph. D. degree in 2011.
She is an Associate Professor at Central South
University, Changsha, China. Her main research
interests include advanced control of power electronics, stability analysis of
time-delay system and power systems, and robust control of FACTS with
wide-area signals.

Canbing Li (M’06–SM’13) was born in Hunan

Province, China, in 1979. He received the B.E. and
Ph.D. degrees from Tsinghua University, Beijing,
China, in 2001 and 2006, respectively, both in elec-
trical engineering.
He is currently a Professor with Hunan University,
Changsha, China.

1949-3053 (c) 2017 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.