See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/344400820

On Grid-Interactive Smart Inverters: Features and Advancements

Article in IEEE Access · September 2020

DOI: 10.1109/ACCESS.2020.3020965

CITATIONS READS

69 1,789

2 authors:

Behrooz Mirafzal Aswad Adib

Kansas State University Kansas State University
134 PUBLICATIONS 3,147 CITATIONS 19 PUBLICATIONS 481 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Gallium Nitride (GaN) devices-based Converters View project

Transient in Motor Drives View project

All content following this page was uploaded by Behrooz Mirafzal on 28 September 2020.

The user has requested enhancement of the downloaded file.

Received August 7, 2020, accepted August 28, 2020, date of publication September 1, 2020, date of current version September 15, 2020.
Digital Object Identifier 10.1109/ACCESS.2020.3020965

On Grid-Interactive Smart Inverters: Features

and Advancements
BEHROOZ MIRAFZAL 1, (Senior Member, IEEE), AND ASWAD ADIB 2, (Member, IEEE)
1 Department of Electrical and Computer Engineering, Kansas State University, Manhattan, KS 66506, USA
2 Oak Ridge National Laboratory, Oak Ridge, TN 37830, USA

Corresponding author: Behrooz Mirafzal ([email protected])

This work was supported in part by the Department of Energy, Office of Energy Efficiency and Renewable Energy (EERE), Solar Energy
Technologies Office, under Award DE-EE0008767, and in part by the Kansas State University Open Access Publishing Fund.

ABSTRACT Traditionally, a grid-interactive inverter providing ancillary services is called a smart inverter.
However, broader features will be required for the next generation of inverters that can be categorized as
self-governing, self-adapting, self-security, and self-healing. For grid-interactive inverters, the self-governing
feature can be identified as the capability of inverters to operate in grid-following and grid-forming control
modes, where the self-adapting is referred to as more flexibility realized by adaptive controllers for stable
dynamics of inverters under various grid conditions. Moreover, for supervisory control and economic
dispatch in a grid with high-penetration of inverter-based power generators, a minimum communication
might be necessary, but it can place grid-interactive inverters in danger of being hacked when self-security
becomes essential to identify malicious setpoints. Furthermore, the self-healing is defined as fault-tolerance
and stress reduction under abnormal conditions. It suggests that after realizing these features, an inverter is
called a smart inverter. In this paper, the advancements toward achieving these features for grid-interactive
inverters are reviewed.

INDEX TERMS Smart inverters, self-security, self-adapting, self-governing, self-healing, cyberattacks.

I. INTRODUCTION
The energy infrastructure is rapidly changing as more sus-
tainable energy resources such as photovoltaic arrays, wind
turbines, and energy storage systems are distributed within
the grid as power generation units. The power generation
from these intermittent dc and ac sources requires solid-state
converters as the interface between the energy resources and
the power grid. These converters typically contain multiple
stages, including an inverter as the grid-side stage. How-
ever, these multi-stage converters are simply known as grid-
interactive inverters, and decentralized generation units are
called distributed generators (DGs). The high-penetration of
DGs provides more flexibility for power systems, where
inverters play the most significant role in DGs. An inverter
with the capability to make proactive and autonomous deci-
sions based on local measurements and external data can
be defined as a smart inverter. Fig. 1 illustrates the desired
features of a smart inverter. For example, inverters can be pro-
grammed to provide ancillary services for power grids under
abnormal conditions to improve power quality. This mode
of operation for inverters is referred to as grid-supporting FIGURE 1. Smart inverters connected to the power grid and a cyber
mode [1], [2]. Inverters can also operate in grid-forming mode network with the features beyond the P and Q grid-feeding duty.

The associate editor coordinating the review of this manuscript and to form an islanded microgrid following a blackout caused
approving it for publication was Ning Kang . by natural disasters, etc. [3]–[8]. Furthermore, inverters play

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
160526 VOLUME 8, 2020
B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

TABLE 1. A classification to define different modes of operation for grid-interactive inverters.

a significant role in making the capability of clustering a

power grid, known as a grid of microgrids or networked
microgrids [9].
The high-penetration of DGs has many benefits but
brings new security, stability, and reliability challenges in
power systems. For example, although decentralized control
schemes, e.g., droop controllers, can address the power-
sharing between DGs, the energy management and eco-
nomic dispatch between DGs cannot be achieved optimally
without a form of the supervisory control scheme, which
needs a type of communication between a utility operator
and inverters [10]–[13]. Thus, a smart inverter becomes a
cyber-physical system that includes both physical parts, e.g.
modules, sensors, and processors, and data packet communi-
cations. The inverters connected to communication or cyber
network could be in danger of being hacked, thus jeopardiz-
ing the security of the inverters. Furthermore, the inherent
zero inertia feature of inverter-based DGs can result in low-
inertia microgrids, making the gird vulnerable to sudden
disturbances and leading to instability. Inverters can also fail FIGURE 2. Typical operating region of a grid-interactive inverter in a
distribution grid with a relatively high ZTh and R/X ratio.
quickly because of internal switch faults, risking the system
reliability. Therefore, self-security, self-adapting, and self-
healing features are also the desired characteristics of a smart
A. GRID-SUPPORTING MODE
inverter besides the self-governing feature. Finally, for the
proper operation of an inverter with different modes of oper- Inverters in power grids encounter various abnormal con-
ation, the inverter should be capable of seamlessly switched ditions. The most commonly occurring abnormalities are
between modes of operation. In other words, smart inverters symmetrical and asymmetrical voltage sags. Inverters are
should be able to self-govern as communicating with agents required by the utilities to stay connected to the system under
or an operator in a supervisory structure. The communication voltage sags, which is referred to as low voltage ride-through
or cyber network provides access to information beyond the (LVRT). Smart inverters should have the capability to detect
available data from their local sensors. The awareness about such grid abnormalities, remain connected to the grid, and
the situation of other DGs, smart meters, and forecasted data consequently provide support to the grid, in the form of
would allow a smart inverter to project anomalies and make reactive power support and negative-sequence compensation.
proactive decisions. In the following subsections, the grid supporting feature
In the following, each section of this paper discusses one of is discussed separately for symmetrical and asymmetrical
the smart inverter features, see Fig. 1. Also, recent advance- anomalies.
ments and existing technical challenges regarding each fea-
ture are briefly presented in the following sections. 1) GRID-SUPPORTING AND SYMMETRICAL ANOMALIES
Inverters should automatically reduce their active power
based on the peak current limit, as shown in Fig. 2, when pro-
II. SELF-GOVERNING FEATURE viding reactive power support [14]–[16]. As the penetration
The self-governing feature is defined as the capability of of inverters is increasing, a sudden loss of power from all the
supporting the grid by providing autonomous ancillary ser- inverter-based DGs could result in a more significant issue
vices or forming microgrids and networked microgrids by than the voltage sag itself. Recent standards, e.g., IEEE Std.
regulating the grid voltage and frequency. These modes of 1547-2018 [16], require inverters to remain connected to the
operation for grid-tied inverters are in contrast with just grid- grid under voltage sags and provide reactive power. However,
feeding, in which the maximum available power at unity if the voltage sag persists for too long, the inverter should
power factor is injected to the grid, see Table 1. then be isolated from the grid [16]. The active power injected

VOLUME 8, 2020 160527

 B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

FIGURE 3. Control block diagram of a PQ-controlled grid following (grid-feeding/grid-supporting) inverter.

into the grid should be adjusted as the reactive power is P2 +Q2 = Smax2 , and thus, losing part of the normal operating

increased to avoid violating the maximum current limit of the region of an inverter in a weak grid. Being inside Smax circle,
inverter, which would otherwise trip the inverter. Along with but outside of the second circle may not necessarily lead to
the maximum current limit, shown as a circle with the radius instability unless the inverter feeds a weak grid when the
of Smax in Fig. 2, the amount of active and reactive power current harmonics can significantly distort the PCC voltage.
accessible for the inverter at a specific time depends on the Fig. 3 shows a grid-supporting inverter, which extracts the
available dc-bus voltage, Vdc , the modulation index, m, and maximum available power, PMPP , of an intermittent source,
the impedance between the inverter and the grid [15], [17], e.g. PV arrays, and feeds the grid. Voltage source invert-
[18]. The second circle with the radius of mVdc Vg / |ZTh | is ers (VSIs) are commonly employed in DG units, whereas
also shown in Fig. 2. To demonstrate this second circle, one current source inverters (CSIs) can also be applied. Unlike a
can write the power transferred from the inverter to the grid VSI that is a buck (step-down) converter, i.e. Vdc > VLL,rms ,
as follows: a CSI is a boost converter [19]–[21], and thus can be fed by a
2 parallel connection of small dc sources, enhancing the relia-
km Vdc Vg e−jδ − Vg
S = P + jQ = ∗ (1) bility and availability of DG units. Regardless of the inverter
ZTh topology, the maximum power capacity of the inverter must
where, Vdc is the dc-bus voltage, km for linear modulation of be curtailed to provide a margin, as demonstrated in Fig. 2,
three-phase inverters is 0.612m for sinusoidal-pulse-width- for injecting reactive power, and subsequently, a capability
modulation (SPWM), and 0.707m for space-vector PWM to regulate the voltage at PCC, i.e. vpcc . As shown in Fig. 3,
(SVPWM), when 0 < m ≤ 1, km Vdc is the line-line rms volt- the grid-supporting feature can be implemented by adjusting
age at the inverter terminals, Vg is the line-line rms voltage of the active and reactive power setpoints using 1P and 1Q
the grid, δ is the angle between the inverter and the grid, and adjustments through a conventional or hybrid droop control
∗ = |Z | e−jθz is the complex conjugate of the Thevenin
ZTh scheme as given below
Th
equivalent impedance of the grid seen by the inverter, which (
P∗ = PMPP − Pc + 1P
includes the filter effect. From (1), one can simply write the (3)
following equation. Q∗ = Qref + 1Q
2
!2 2
!2 where, 1Q can be formulated as
Vg cos (θz ) Vg sin (θz )
P+ + Q+ (
|ZTh | |ZTh | (−1/mQ )1V |1V | > Vth
1Q = (4)
!2 0 |1V | ≤ Vth
km Vdc Vg
= (2) where, Vth is the threshold voltage determined by grid codes,
|ZTh |
e.g. IEEE Std. 1547, and mQ is the droop control coeffi-
where, Eq. (2) demonstrates a circle with the radius of cient. Notice, the output of the droop, 1Q, must be limited
2
mVdc Vg / |ZTh | and the center of − Vg /ZTh ∗ . As depicted between 1Qmax(−) and 1Qmax(+) , as one can derive from the
in Fig. 2, an increase in the R/X ratio the center of this normal operating region shown in Fig. 2. The active power
circle is shifted to the left, and a decrease in the available can similarly be tuned to support the grid, particularly in
dc-bus voltage, Vdc , or a low grid voltage, Vg , in which the distribution power grid, in which R/X ratio is relatively
ZTh has a relatively high value in weak grids, shrinks the high. The grid-supporting feature is not limited to the voltage
disk identified by the circle. Any change in these parameters support. Inverters can also operate as harmonic and negative
can result in intersecting this circle with the Smax circle, i.e. sequence compensators in asymmetrical three-phase systems.

160528 VOLUME 8, 2020

B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

to the PWM reference signals to keep the inverter in the linear

modulation region while providing ancillary services such as
negative sequence and harmonic compensations [26].

B. GRID-FORMING MODE
In contrast to a grid-following inverter, a grid-forming
inverter operates as a voltage source, see Fig. 5, and
Table 1. In a power grid, not all inverters need to work
in grid-forming mode [3]–[8]. Generally, inverters powered
by non-intermittent sources such as battery energy storage
and natural gas microturbines are chosen to be grid-forming
FIGURE 4. Block diagram and decision table for negative-sequence
inverters, while the remaining inverters can operate in their
controller in grid-supporting mode. original mode of operation, e.g. grid-feeding mode. Inverters
powered by renewable sources such as PV arrays and wind
turbines can also operate in grid-forming mode, but only
2) GRID-SUPPORTING AND ASYMMETRICAL ANOMALIES
when they are equipped with their own energy storage units.
Asymmetrical anomalies can occur due to asymmetrical A grid-forming inverter must be sized for higher-rated values
faults in the power grid and uneven distribution of single- compared to an equivalent grid-feeding inverter to provide
phase loads in the distribution grid. A smart inverter can sufficient flexibility and a stability margin, particularly in
support the grid in such instances by providing negative- asymmetrical and weak grids.
sequence compensation services on top of the positive- To form islanded microgrids, and possibly form a grid
sequence reactive power support [22]–[26]. of these microgrids, smart inverters must optimally share
As shown in Fig. 4, supplementary q- and d-axis current the total load among themselves and other power gener-
control paths can be inserted to the control scheme for pro- ation units. The power-sharing and the control of voltage
viding negative-sequence compensation to the grid currents, and frequency can be achieved using either centralized or
irrespective of whether the control scheme is formulated decentralized techniques. The main features of centralized
based on stationary or synchronous reference frames [22]. and decentralized techniques are discussed in the following
These paths feed the negative-sequence components of the subsections.
grid current back to the controller, as shown in Fig. 4, with a
zero-reference value to achieve balanced grid currents while
1) CENTRALIZED CONTROL TECHNIQUE
the local load might be asymmetrical. However, negative-
sequence compensation provided in such a way does not com- In the centralized control approach, inverters adopt a com-
pensate for the negative-sequence component of the voltage at munication based secondary control structure [10]–[13], [27],
PCC. Therefore, oscillations in the active and reactive power [28], which can be categorized as master-slave, concentrated,
may still remain, which in turn can cause a ripple in the and distributed control techniques. The master-slave and dis-
dc-bus voltage of the inverter. The ripples extending beyond tributed control techniques are explained below.
the voltage ratings of the capacitor can damage the dc-bus
capacitors. The setpoints of the negative-sequence controller a: MASTER-SLAVE CONTROL TECHNIQUE
can be adjusted to any value other than zero to mitigate some In a master-slave scheme, an inverter operates as a voltage
of the oscillations. In general, the desired values of the con- source to regulate voltage magnitude and frequency while
troller can be set to either mitigate active and reactive power other inverters operate as current sources [29], [30]. The
oscillations or compensate unbalanced grid currents [23], see inverter that regulates the voltage and frequency is called
Fig. 4. More control objectives for the negative-sequence con- master-inverter. Master inverters softly start before slave-
trollers have also been proposed in the literature [24], [25]. inverters feed an islanded microgrid. For small microgrids,
In [24], the setpoints of the negative-sequence controllers are a single grid-forming or master-inverter is sufficient to reg-
designed to reduce the negative-sequence component of the ulate the amplitude and frequency of the voltage at the
voltage at PCC. In [25], the objective of the controllers is main ac bus. However, for a large microgrid, one inverter
to reduce the power oscillation as much as possible while would not be able to handle the load variations of the micro-
also limiting the currents at their rated values. A smart grid, and as such, multiple grid-forming or master-inverters
inverter equipped with a negative-sequence compensation might be required. To synchronize the grid-forming inverters,
feature should be capable of providing such services based a communication link, which will share the phase and fre-
on the appropriate control objective depending on the specific quency of a reference signal, is required [11], [29], [30]. The
situation. Since the inverter references are unbalanced when master-slave technique provides outstanding power-sharing
providing negative-sequence compensation, some of the ref- performance. In the case of tripping the master-inverter, one
erences may reach the overmodulation region even at rated of the slave-inverters needs to be seamlessly switched to
power. A low-frequency common-mode signal can be added operate as the new master-inverter.

VOLUME 8, 2020 160529

 B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

FIGURE 5. Inverters in different modes of operation in grid-following for grid-tied systems, and grid-forming for islanded microgrid systems.

FIGURE 6. A demonstration of power sharing between three inverters with different capacity, Pimax , using P-f droop control technique, these
inverters contribute to the load change, 1P L , based on their capacities or droop coefficients, mPi , as 1P L = 1P1 + 1P2 + 1P3 .

b: DISTRIBUTED CONTROL TECHNIQUE

In a distributed control scheme, each inverter regulates the
voltage and frequency of the output voltage based on the
grid nominal reference values [11], [27], [28], [30], [31], e.g.
240V, 60Hz; however, those are slightly adjusted based on
the average current of all inverters in a microgrid, i.e. for a
microgrid with N inverter the average of output currents is
given by
N FIGURE 7. A basic block diagram for P-f and Q-V droop power sharing.
1 X
iavg = ipcc,k (5)
N
k=1 for power-sharing [11], [31]–[34]. The power-sharing con-
The error of the output current for each inverter with respect cept using P-f droop lines is demonstrated in Fig. 6. The
to the average current is then calculated as iek = iavg − ipcc,k . control scheme still needs a PI voltage control loop. A basic
The error signal is decomposed into dq frame of reference droop control block diagram is shown in Fig. 7. However,
and used to adjust the frequency and voltage controller [11]. the reference voltage information is generated by the droop
These adjustments can be tuned further to share the total load equations. The P-f and Q-V droop equations can be written
based on economic dispatch. Like the droop control scheme as:
(
explained in the next subsection, the frequency and voltage 1f = −mPi 1Pi i = 1, 2, . . . , N
have the flexibility to deviate slightly from their nominal (6)
1V = −mQi 1Qi i = 1, 2, . . . , N
values. The nominal frequency should be restored follow-
ing the power-sharing stage. Many investigations have been where, mPi and mQi are the droop coefficients of the ith
reported on various distributed control methods, for example, inverter, the frequency is given by f ∗ = 1f + fref , and the
a technique in which each inverter needs the data from its desired voltage at PCC is obtained from V ∗ = 1V + Vref .
local sensors and the information of only nearby DG units to Notice, for distribution grids and more resistive circuits, P-V ,
minimize the required communication network [27]. and Q-f droop control schemes have demonstrated better
power-sharing among the inverters. The former droop equa-
2) DECENTRALIZED CONTROL TECHNIQUE tions can still be used in high R/X ratio grids with the addition
Two of the most known decentralized control schemes are of virtual inductance/impedance in the control scheme, which
droop and virtual synchronous generator (VSG) techniques. makes the output impedance of the inverter more induc-
tive [33]. The amount of power each grid-forming inverter
a: DROOP CONTROL TECHNIQUE will share can be controlled through its droop coefficient,
A droop control scheme provides a decentralized approach, which is a function of the rating of the inverter, see Fig. 6.
which is inherited from the parallel operation of synchronous Reactive power-sharing can sometimes be challenging using
generators. The grid-forming inverters equipped by droop droop control because of the unequal voltage drop across
controllers do not necessarily require a communication link line impedances for the grid-forming inverters. An adaptive

160530 VOLUME 8, 2020

B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

FIGURE 8. A basic block diagram for VSG control technique. FIGURE 9. Block diagram of grid resynchronization control.

virtual inductance can be helpful in such cases. Therefore, inverter for providing the energy reserve necessary to supply
the virtual inductance for droop control should have a static virtual inertia [35].
element, which should stay within 0.05-0.15 pu to ensure that Centralized schemes for grid-forming inverter such as the
the equation for power transfer holds [32]. Besides, an adap- master-slave approach might be more suitable for micro-
tive virtual impedance should be incorporated to ensure that grids containing inverters in close proximity such as in smart
reactive power is properly shared among grid-forming invert- buildings, where a communication link between the inverters
ers [33]. Furthermore, a droop control method should achieve can be quickly established. For microgrids spread over a
zero error frequency restoration. distribution network, a decentralized approach might be more
suitable, whereas the droop control or virtual synchronous
generator scheme can be used based on the availability of
b: VIRTUAL SYNCHRONOUS GENERATOR CONTROL
energy storage units.
TECHNIQUE
The virtual synchronous generator (VSG) control technique
III. FLEXIBILITY AND SELF-ADAPTING FEATURE
for grid-interactive inverters mimics the dynamic behavior of
electric machines, see Fig. 8. To fully achieve VSG, a DG Forming an islanded microgrid may place inverters in a
should be equipped by dc-bus energy storage to play the role weak grid condition. Therefore, self-adaptation should be
of the kinetic energy reservoir in the rotating mass of an added into the controller of smart inverters for larger stability
electric machine [34], [35]. The swing equation derived from regions and seamless transitions between modes of operation.
the equation of motion can be expressed as:
A. SEAMLESS TRANSITION AND ISLANDING DETECTION
ωm − ω
 
dωm Power grids can encounter a loss of utility power because of
Pm − P = J ωm +D Sb (7)
dt ωref natural events or physical and cyberattacks. Therefore, invert-
ers should be programmed to detect loss of utility power,
where, J denotes the virtual inertia, D is the damping fac- and consequently, shape islanded microgrids. The terminal
tor, ωm , ω, and ωref are the virtual rotor angular frequency, voltage of inverters is measured by their local controllers.
the angular frequency of the voltage at PCC, and nominal Therefore, the simplest way to detect loss of grid is to deter-
angular frequency, respectively. Also, Pm , P, and Sb are the mine whether the measured voltage and frequency are outside
virtual shaft power determined by the droop (governor) equa- the normal range of operation defined in the IEEE Std. 1547-
tion, the active power injected to the grid measured at PCC, 2018 [16]. The time interval during which the voltage falls
and the power rating of the inverter, respectively. outside the normal range must be considered, according to
The inertia of inverter-based DGs is inherently zero unless IEEE Std. 1547-2018, before performing any reaction to
some virtual inertia has been implemented into the con- avoid false detection. The inverter can be separated from
trol schemes. Therefore, microgrids with a high number of the grid using a static transfer switch (STS). When a single
inverter-based DGs can be susceptible to sudden load changes inverter is operating in an islanded mode, reconnection to the
and disturbances. Particularly for grid-forming inverters in grid is simple since the voltage buildup at the other end of the
droop mode, the under-frequency, and over-frequency relays STS can be detected, and the inverter can be synchronized to
can be triggered under sudden changes [34]. The virtual the grid before reconnecting. To avoid high current transients,
synchronous generator method can be implemented for bet- the difference in voltage amplitude and phase across the
ter dynamic performances by adding virtual inertia to the STS can be fed to two PI controllers, which would then
inverter. Similar Q-V and P-f equations, as given in (6), can generate compensation signals to be added to the inverter
be combined with (7) to form a virtual synchronous generator voltage amplitude and phase for slowly resynchronizing to
in which the P-f droop equation is the governor of the virtual the grid, as shown in Fig. 9 [30]. However, the STS could
synchronous generator, as shown in Fig. 8. Furthermore, be far away from the inverter, and as such, the synchro-
an energy storage element can be added to the dc-bus of the nizing information would not be readily available. In such

VOLUME 8, 2020 160531

 B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

cases, the microgrid has to be equipped with some sort of

supervisory controller, which can provide the synchronizing
information through low bandwidth communication signals
[30]. Alternatively, there are methods that utilize only the
inverter-side current and voltage measurements with no con-
trol over the grid-side of STS. These methods typically
require a quick detection of the grid phase angle to minimize
the transient overcurrent [36], [37].

B. SUPERVISORY CONTROL USING FORECASTING DATA

FIGURE 10. A model-reference approach to examine and determine
Besides the economic dispatch, a supervisory or tertiary cyberattacks versus healthy supervisory control commands from the
controller can enable the interconnection of microgrids to utility.
form a grid of microgrid through sharing synchronizing data
[9]–[13], [38], [39]. Additionally, some issues arising from controller. However, it requires additional circuitry contribut-
the intermittent nature of renewable energy sources can ing to additional size and cost. In [46], the gain of the voltage
be mitigated with a supervisory structure that has access feedforward path has been updated adaptively to enhance
to weather forecasting data. On a particularly windy day, stability in weak grids. However, the technique presented in
the wind speed limit of the wind turbines might be violated, [46] required grid impedance estimation, implemented using
which will cause the wind turbines to shut down. Similarly, a band-pass filter. In summary, to make smart inverters robust
the output of a PV array can suddenly drop following the pass- against weak grids, adaptive elements need to be introduced
ing of a large cloud. As the penetration of renewable energy in the control scheme. While the conventional controllers
sources is increasing, the loss of a wind farm or solar farm with fixed parameters are commonly used, adding adaptive
could cause frequency and voltage oscillations. Alternately, feedforward and virtual impedance schemes seem to be the
if the weather forecast is known beforehand, the solar and most promising solutions to realize the adapting feature for
wind farms can be gradually powered down while the remain- smart inverters.
ing sources, particularly high inertia synchronous generators,
can be slowly powered up to avoid any under-frequency trips. IV. COMMUNICATION AND SELF-SECURITY FEATURE
The performance of smart inverters in both grid-forming and
C. ADAPTIVE STABILIZERS grid-following modes can be improved with access to exter-
Weak grids make inverters susceptible to voltage devia- nal data. However, data packet transmissions between smart
tions, while a severe weak grid case can make the inverter inverters and the utility operator through a communication
unstable, which will result in sudden inverter disconnec- network can make inverters subject to cyberattacks and unin-
tion. The sudden loss of power from an inverter in islanded tentional human mistakes [47], [48]. As revealed in Fig. 10,
microgrids could cause a cascaded failure event in a newly this challenge can be addressed through implementing a ref-
formed microgrid after the loss of utility power. The current erence system (model) to distinguish a malicious setpoint
control loop bandwidth, voltage feedforward path, phase- from a regular setpoint receiving from the power utility.
locked loop (PLL), and filter parameters are the elements Secure communication can be achieved using known message
that can be appropriately designed for improved dynamic authentication code (MAC) methods, to confirm if a setpoint
performances [37], [40]–[43]. came from the utility and has not been altered. Nonetheless,
As discussed earlier, the grid parameters play a significant the utility computer can be hacked, and manipulated setpoints
role in shrinking the normal operating region of inverters can be sent with secure tags. One way to form a self-secure
in weak grids. Thus, including an adaptive element into inverter is to identify the normal operating region, see Fig. 2,
the controller of smart inverters is advancing the inverter and used it along with the grid codes as the knowledge-base.
flexibility [44]–[46]. In [44], an adaptive control scheme to If a hacker manipulates the inverter setpoints, the inverter can
enhance the stability of inverters is presented based on the first examine the new setpoints using the reference model, and
online estimation of grid impedance. For larger values of grid then, may refuse to engage them if the projected output falls
impedance, the PLL bandwidth is lowered to keep the inverter outside of the safe operating region of the inverter. Based on
in the stable region. However, the PLL bandwidth had to be various types of cyberattacks [49], [50], different and more
lowered considerably to ensure stability, thus introducing a sophisticated methods can be developed.
tradeoff between stability and dynamic performance. In [45], A communication network for grid-interactive inverters
an active damper is added to the system, which essentially must be secure and scalable while providing low latency, high
introduces an additional resistive term in the inverter circuit range, and adequate data rate [10]. The IEC 61850 standard
that can be varied adaptively to make the inverter more robust provides a platform for advanced interoperability of intel-
against changes in a grid condition. The active damper is ligent electronic devices (IEDs), e.g. smart inverters, from
realized using an additional low-power single-phase inverter, different vendors [12], [13]. In IEC 61850, the data model
which ensures stability without adding any complexity to the can be mapped to multiple protocols, e.g. GOOSE, MMS,

160532 VOLUME 8, 2020

B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

FIGURE 12. Timeline of transient and dynamic phenomena in power grids

FIGURE 11. A basic implementation for a communication network needed and maximum latency requirements for transmitting protection and
between the utility operator, smart inverters, and measurement units. control data packets based on the IEC 61850 protocol.

trigger a cascaded event similar to the weak grid case that

and SMV. The generic-object-oriented-substation-events may collapse the entire system.
(GOOSE) protocol is for exchanging information between Unlike the other features considered for smart invert-
IEDs in a publisher/subscriber arrangement, whereas the ers, real-time fault diagnosis, and fault-tolerance techniques
manufacturing- message-specification (MMS) protocol is for inverters in different applications such as in motor
for transferring real-time data and supervisory control data drives, and electric powertrains can be applied for grid-
between IEDs and the utility operator in a client/server for- tied inverters [54], as long as their implementation is not
mat, and the sample-measured-values (SMV) protocol is for application-specific.
transmitting digitized signals from measurement units, e.g. Fault-tolerant inverters can be broadly categorized as non-
sensors, to IEDs, see Fig. 11. Notice, the processing and redundant and redundant techniques [55]. Following the
end-to-end data packet transmissions have inherent delays, detection of an incipient fault, inverters using non-redundant
which will confine the effectiveness of a centralized control techniques have to switch to a different control scheme to
scheme on suppressing transient and fast dynamic phenom- continue operation with a fewer number of active devices. The
ena [10], [12], [13], [51]–[53], see Fig. 12. One may observe main components, e.g., semiconductor devices and dc-bus
that a cyberattack can be planned by altering the end-to-end capacitors, also have to be overrated for non-redundant tech-
packet delays between smart inverters, sensors, and the utility niques. Even if an inverter remains operational after isolating
for a specific event, making the overall system unstable. an internal fault, the resulting grid currents can be severely
A communication network can be wired, e.g. optical fiber, asymmetric with non-redundant techniques.
power line communication (PLC), etc., wireless, e.g. cellular, Redundant techniques for two-level inverters require an
Wi-Fi, etc., or a combination of both [10], [12], [13]. The additional leg with two semiconductor devices. For two-level
wired communication methods are typically more immune to inverters equipped with an auxiliary leg, the faulty leg needs
electromagnetic interferences (EMI), but they are less scal- to be isolated, and the auxiliary leg has to connect to the
able in comparison with the wireless techniques. Using power circuit appropriately. In redundant fault-tolerant inverters,
lines for data communication is the technology that has a long each switch is equipped with a fast-response overcurrent fuse
history in relay and protection systems and provides a low- to isolate the device in case of a short-circuit fault. Once
cost solution, but it has the lowest data rate compared to other the device is isolated, that branch of the circuit will act
wired communication technologies. Also, inverters may lose similarly to the case of an open-circuit fault. Furthermore,
their access to the external data in islanded mode using power an appropriate open-circuit fault detection technique should
lines for data communication. A wireless communication detect the location of a fault. Depending on the location of
network with mesh topology is more fault tolerance, but due a fault, the proper connecting switch should be closed to
to the routing process, the actual data rate may significantly introduce a device from the auxiliary leg into the circuit.
be reduced to an unacceptable level. To avoid complicated Low-speed semiconductor switches are typically used as the
methods and provide high scalability, sparse communication connecting switches, as the connecting switches commutate
technique has been proposed as a solution in which inverters at the fundamental frequency.
need to communicate only with their nearby smart devices Although two-level inverters are commonly used for grid-
and inverters [27], [28]. tied applications, modular multilevel inverters are inherently
more tolerable of internal faults than two-level inverters [54],
V. FAULT DIAGNOSIS AND SELF-HEALING FEATURE [56]–[62]. However, they are normally used for high and
Incipient fault diagnosis and self-healing of inverters is a cru- medium voltage applications. If the self-healing feature is
cial feature. It becomes even more critical for inverters oper- desired, multilevel inverters should be recommended. One of
ating in islanded microgrids and grids with high penetration the best topologies for realizing a fault-tolerant inverter is the
of inverter-based DGs. In islanded microgrids, the total load cascaded h-bridge (CHB) inverter, where the inverter neutral
of the system is shared among the inverters. As a result, if an point is not grounded. A discrete method, called fundamen-
inverter gets suddenly disconnected due to a fault, it could tal phase-shift compensation [59], [60], is demonstrated in

VOLUME 8, 2020 160533

 B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

decentralized control schemes for grid-forming operations

have been studied for islanded mode. The loss of utility power
detection and seamless resynchronization to the grid has been
referred to as the grid-adapting feature. Moreover, adaptive
and virtual elements have been recognized as valuable fea-
tures for inverters in weak grids. The importance of engaging
communication networks to control grid-tied inverters has
been discussed. The need for device-level security against
cyberattacks has been identified and referred to as the self-
security feature. Furthermore, the communication require-
ments and the latency problem have been briefly discussed.
FIGURE 13. Phasor diagrams in the prefault and a postfault scenario after Lastly, the ability to detect internal faults and perform nec-
isolating two cells of Phase-C of an 11-level CHB inverter when the
phase-shift compensation technique has been implemented. essary corrective actions have been identified as a desired
feature for smart inverters, preventing the inverter from a
TABLE 2. Corrective angles after bypassing one and two faulty cells in a
sudden shutdown. These features intensify the role of a grid-
11-level CHB inverter. tied inverter from being just a reactionary to a proactive
device, and all are the hallmarks of the next-generation smart
inverters.

REFERENCES
[1] J. Hu, Z. Li, J. Zhu, and J. M. Guerrero, ‘‘Voltage stabilization: A critical
step toward high photovoltaic penetration,’’ IEEE Ind. Electron. Mag.,
vol. 13, no. 2, pp. 17–30, Jun. 2019.
[2] S. A. Khajehoddin, M. Karimi-Ghartemani, and M. Ebrahimi, ‘‘Grid-
supporting inverters with improved dynamics,’’ IEEE Trans. Ind. Electron.,
Fig. 13, in which the voltage magnitude is degraded, but vol. 66, no. 5, pp. 3655–3667, May 2019.
the three-phase system remains symmetrical after bypassing [3] P. J. Hart, R. H. Lasseter, and T. M. Jahns, ‘‘Coherency identification and
faulty cells, with a floating neutral point. In Table 2, two aggregation in grid-forming droop-controlled inverter networks,’’ IEEE
Trans. Ind. Appl., vol. 55, no. 3, pp. 2219–2231, May 2019.
fault scenarios and their corresponding corrective angles are [4] R. H. Lasseter, Z. Chen, and D. Pattabiraman, ‘‘Grid-forming inverters:
given in Table 2. Notice, the CHB inverter in this table is A critical asset for the power grid,’’ IEEE J. Emerg. Sel. Topics Power
an 11-level inverter, in which
√ the line-line prefault voltage
Electron., vol. 8, no. 2, pp. 925–935, Jun. 2020.
magnitude is equal to 5 3 p.u., assuming the voltage of [5] B. Mahamedi and J. E. Fletcher, ‘‘The equivalent models of grid-forming
inverters in the sequence domain for the steady-state analysis of power
each cell is 1 p.u. In [61], a continuous self-healing technique systems,’’ IEEE Trans. Power Syst., vol. 35, no. 4, pp. 2876–2887,
has been presented. This technique enables CHB inverters to Jul. 2020.
provide a symmetrical ac voltage from a set of asymmetrical [6] Y. Du, X. Lu, H. Tu, J. Wang, and S. Lukic, ‘‘Dynamic microgrids with
self-organized grid-forming inverters in unbalanced distribution feeders,’’
time-variant input dc source voltages using an adaptive PWM IEEE J. Emerg. Sel. Topics Power Electron., vol. 8, no. 2, pp. 1097–1107,
technique. Two alternative methods for implementing adap- Jun. 2020.
tive SPWM have been proposed in [61]. The first method, [7] X. Huang, K. Wang, J. Qiu, L. Hang, G. Li, and X. Wang, ‘‘Decentralized
called the sensor-per-source (SPS) algorithm, requires all cell control of multi-parallel grid-forming DGs in islanded microgrids for
enhanced transient performance,’’ IEEE Access, vol. 7, pp. 17958–17968,
input voltages to be monitored in real-time while the sec- 2019.
ond method, called the sensor-per-leg (SPL) algorithm, only [8] S. Yazdani, M. Ferdowsi, M. Davari, and P. Shamsi, ‘‘Advanced current-
requires real-time monitoring of the leg’s line-to-neutral out- limiting and power-sharing control in a PV-based grid-forming inverter
under unbalanced grid conditions,’’ IEEE J. Emerg. Sel. Topics Power
put voltage. More comprehensive self-healing techniques and Electron., vol. 8, no. 2, pp. 1084–1096, Jun. 2020.
corrective actions can be implemented for modular battery- [9] Z. Li, M. Shahidehpour, F. Aminifar, A. Alabdulwahab, and Y. Al-Turki,
powered grid-forming inverters to enhance the availability of ‘‘Networked microgrids for enhancing the power system resilience,’’ Proc.
smart inverters. IEEE, vol. 105, no. 7, pp. 1289–1310, Jul. 2017.
[10] B. Arbab-Zavar, E. Palacios-Garcia, J. Vasquez, and J. Guerrero, ‘‘Smart
inverters for microgrid applications: A review,’’ Energies, vol. 12, no. 5,
VI. CONCLUSION p. 840, Mar. 2019.
The desired features of smart inverters have been catego- [11] H. Han, X. Hou, J. Yang, J. Wu, M. Su, and J. M. Guerrero, ‘‘Review of
power sharing control strategies for islanding operation of AC microgrids,’’
rized as self-governing, self-adapting, self-security, and self- IEEE Trans. Smart Grid, vol. 7, no. 1, pp. 200–215, Jan. 2016.
healing in this paper. The self-governing role of inverters [12] I. Serban, S. Céspedes, C. Marinescu, C. A. Azurdia-Meza, J. S. Gomez,
has been defined as providing voltage support and ancillary and D. S. Hueichapan, ‘‘Communication requirements in microgrids: A
services for the power grid, ensuring the grid codes and stan- practical survey,’’ IEEE Access, vol. 8, pp. 47694–47712, 2020.
[13] S. Kumar, S. Islam, and A. Jolfaei, ‘‘Microgrid communications—
dards. Also, constraints of providing ancillary services under Protocols and standards,’’ in Variability, Scalability and Stability of
unbalanced anomalies have been determined. Furthermore, Microgrids. Jul. 2019, pp. 291–326. [Online]. Available: https://digital-
the advantages and disadvantages of both centralized and library.theiet.org/content/books/10.1049/pbpo139e_ch9

160534 VOLUME 8, 2020

B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

[14] T. Stetz, F. Marten, and M. Braun, ‘‘Improved low voltage grid-integration [35] X. Meng, J. Liu, and Z. Liu, ‘‘A generalized droop control for grid-
of photovoltaic systems in germany,’’ IEEE Trans. Sustain. Energy, vol. 4, supporting inverter based on comparison between traditional droop control
no. 2, pp. 534–542, Apr. 2013. and virtual synchronous generator control,’’ IEEE Trans. Power Electron.,
[15] J. Lamb and B. Mirafzal, ‘‘Grid-interactive cascaded H-bridge multilevel vol. 34, no. 6, pp. 5416–5438, Jun. 2019.
converter PQ plane operating region analysis,’’ IEEE Trans. Ind. Appl., [36] D. S. Ochs, B. Mirafzal, and P. Sotoodeh, ‘‘A method of seamless tran-
vol. 53, no. 6, pp. 5744–5752, Nov. 2017. sitions between grid-tied and stand-alone modes of operation for utility-
[16] IEEE Standard for Interconnection and Interoperability of Distributed interactive three-phase inverters,’’ IEEE Trans. Ind. Appl., vol. 50, no. 3,
Energy Resources With Associated Electric Power Systems Interfaces, pp. 1934–1941, May/Jun. 2014.
IEEE Standard 1547-2018, 2018. [37] F. Sadeque, J. Benzaquen, A. Adib, and B. Mirafzal, ‘‘Direct phase-angle
[17] J. M. Guerrero, J. Matas, L. G. de Vicuna, M. Castilla, and J. Miret, detection for three-phase inverters in asymmetrical power grids,’’ IEEE
‘‘Decentralized control for parallel operation of distributed generation J. Emerg. Sel. Topics Power Electron., early access, Mar. 2, 2020, doi:
inverters using resistive output impedance,’’ IEEE Trans. Ind. Electron., 10.1109/JESTPE.2020.2977398.
vol. 54, no. 2, pp. 994–1004, Apr. 2007. [38] V. Madani, R. Das, F. Aminifar, J. McDonald, S. S. Venkata, D. Novosel,
[18] J. Lamb, B. Mirafzal, and F. Blaabjerg, ‘‘PWM common mode reference A. Bose, and M. Shahidehpour, ‘‘Distribution automation strategies chal-
generation for maximizing the linear modulation region of CHB convert- lenges and opportunities in a changing landscape,’’ IEEE Trans. Smart
ers in islanded microgrids,’’ IEEE Trans. Ind. Electron., vol. 65, no. 7, Grid, vol. 6, no. 4, pp. 2157–2165, Jul. 2015.
pp. 5250–5259, Jul. 2018. [39] M. G. Kashani, M. Mobarrez, and S. Bhattacharya, ‘‘Smart inverter volt-
[19] A. Singh and B. Mirafzal, ‘‘An efficient grid-connected three-phase single- watt control design in high PV-penetrated distribution systems,’’ IEEE
stage boost current source inverter,’’ IEEE Power Energy Technol. Syst. J., Trans. Ind. Appl., vol. 55, no. 2, pp. 1147–1156, Mar./Apr. 2019.
vol. 6, no. 3, pp. 142–151, Sep. 2019. [40] A. Adib, B. Mirafzal, X. Wang, and F. Blaabjerg, ‘‘On stability of voltage
[20] A. Singh, A. K. Kaviani, and B. Mirafzal, ‘‘On dynamic models and source inverters in weak grids,’’ IEEE Access, vol. 6, pp. 4427–4439, 2018.
stability analysis of three-phase phasor PWM-based CSI for stand-alone [41] S. Silwal, S. Taghizadeh, M. Karimi-Ghartemani, M. J. Hossain, and
applications,’’ IEEE Trans. Ind. Electron., vol. 62, no. 5, pp. 2698–2707, M. Davari, ‘‘An enhanced control system for single-phase inverters inter-
May 2015. faced with weak and distorted grids,’’ IEEE Trans. Power Electron., vol. 34,
[21] A. Singh, J. Benzaquen, and B. Mirafzal, ‘‘Current source Generator– no. 12, pp. 12538–12551, Dec. 2019.
Converter topology for direct-drive wind turbines,’’ IEEE Trans. Ind. Appl., [42] X. Li, J. Fang, Y. Tang, X. Wu, and Y. Geng, ‘‘Capacitor-voltage feed-
vol. 54, no. 2, pp. 1663–1670, Apr. 2018. forward with full delay compensation to improve weak grids adaptability
[22] A. Camacho, M. Castilla, J. Miret, A. Borrell, and L. G. de Vicuña, ‘‘Active of LCL-filtered grid-connected converters for distributed generation sys-
and reactive power strategies with peak current limitation for distributed tems,’’ IEEE Trans. Power Electron., vol. 33, no. 1, pp. 749–764, Jan. 2018.
generation inverters during unbalanced grid faults,’’ IEEE Trans. Ind. [43] A. Adib and B. Mirafzal, ‘‘Virtual inductance for stable operation of grid-
Electron., vol. 62, no. 3, pp. 1515–1525, Mar. 2015. interactive voltage source inverters,’’ IEEE Trans. Ind. Electron., vol. 66,
no. 8, pp. 6002–6011, Aug. 2019.
[23] R. Kabiri, D. G. Holmes, and B. P. McGrath, ‘‘Control of active and
[44] M. Cespedes and J. Sun, ‘‘Adaptive control of grid-connected inverters
reactive power ripple to mitigate unbalanced grid voltages,’’ IEEE Trans.
based on online grid impedance measurements,’’ IEEE Trans. Sustain.
Ind. Appl., vol. 52, no. 2, pp. 1660–1668, Mar. 2016.
Energy, vol. 5, no. 2, pp. 516–523, Apr. 2014.
[24] F. Nejabatkhah, Y. W. Li, and B. Wu, ‘‘Control strategies of three-phase dis-
[45] L. Jia, X. Ruan, W. Zhao, Z. Lin, and X. Wang, ‘‘An adaptive active damper
tributed generation inverters for grid unbalanced voltage compensation,’’
for improving the stability of grid-connected inverters under weak grid,’’
IEEE Trans. Power Electron., vol. 31, no. 7, pp. 5228–5241, Jul. 2016.
IEEE Trans. Power Electron., vol. 33, no. 11, pp. 9561–9574, Nov. 2018.
[25] M. M. Shabestary and Y. A.-R.-I. Mohamed, ‘‘Analytical expressions
[46] J. Xu, S. Xie, Q. Qian, and B. Zhang, ‘‘Adaptive feedforward algorithm
for multiobjective optimization of converter-based DG operation under
without grid impedance estimation for inverters to suppress grid current
unbalanced grid conditions,’’ IEEE Trans. Power Electron., vol. 32, no. 9,
instabilities and harmonics due to grid impedance and grid voltage distor-
pp. 7284–7296, Sep. 2017.
tion,’’ IEEE Trans. Ind. Electron., vol. 64, no. 9, pp. 7574–7586, Sep. 2017.
[26] A. Adib, J. Lamb, and B. Mirafzal, ‘‘Ancillary services via VSIs in micro- [47] T. S. Ustun, ‘‘Cybersecurity vulnerabilities of smart Inverters and their
grids with maximum DC-bus voltage utilization,’’ IEEE Trans. Ind. Appl., impacts on power system operation,’’ in Proc. Int. Conf. Power Elec-
vol. 55, no. 1, pp. 648–658, Jan. 2019. tron., Control Autom. (ICPECA), New Delhi, India, 2019, pp. 1–4, doi:
[27] X. Wu, C. Shen, and R. Iravani, ‘‘A distributed, cooperative frequency and 10.1109/ICPECA47973.2019.8975537.
voltage control for microgrids,’’ IEEE Trans. Smart Grid, vol. 9, no. 4, [48] Y. He, G. J. Mendis, and J. Wei, ‘‘Real-time detection of false data injection
pp. 2764–2776, Jul. 2018. attacks in smart grid: A deep learning-based intelligent mechanism,’’ IEEE
[28] A. Bidram, A. Davoudi, and F. L. Lewis, ‘‘A multiobjective distributed con- Trans. Smart Grid, vol. 8, no. 5, pp. 2505–2516, Sep. 2017.
trol framework for islanded AC microgrids,’’ IEEE Trans. Ind. Informat., [49] S. Tan, J. M. Guerrero, P. Xie, R. Han, and J. C. Vasquez, ‘‘Brief survey on
vol. 10, no. 3, pp. 1785–1798, Aug. 2014. attack detection methods for cyber-physical systems,’’ IEEE Syst. J., early
[29] T. Caldognetto and P. Tenti, ‘‘Microgrids operation based on master–slave access, May 20, 2020, doi: 10.1109/JSYST.2020.2991258.
cooperative control,’’ IEEE J. Emerg. Sel. Topics Power Electron., vol. 2, [50] D. Ding, Q.-L. Han, Y. Xiang, X. Ge, and X.-M. Zhang, ‘‘A survey on secu-
no. 4, pp. 1081–1088, Dec. 2014. rity control and attack detection for industrial cyber-physical systems,’’
[30] X. Hou, Y. Sun, J. Lu, X. Zhang, L. H. Koh, M. Su, and J. M. Guerrero, Neurocomputing, vol. 275, pp. 1674–1683, Jan. 2018.
‘‘Distributed hierarchical control of AC microgrid operating in grid- [51] Y. Gao and Q. Ai, ‘‘A distributed coordinated economic droop control
connected, islanded and their transition modes,’’ IEEE Access, vol. 6, scheme for islanded AC microgrid considering communication system,’’
pp. 77388–77401, 2018. Electr. Power Syst. Res., vol. 160, pp. 109–118, Jul. 2018.
[31] Q. Zhou, Z. Li, Q. Wu, and M. Shahidehpour, ‘‘Two-stage load [52] S. Liu, X. Wang, and P. X. Liu, ‘‘Impact of communication delays on
shedding for secondary control in hierarchical operation of islanded secondary frequency control in an islanded microgrid,’’ IEEE Trans. Ind.
microgrids,’’ IEEE Trans. Smart Grid, vol. 10, no. 3, pp. 3103–3111, Electron., vol. 62, no. 4, pp. 2021–2031, Apr. 2015.
May 2019. [53] J. S. Gomez, D. Saez, J. W. Simpson-Porco, and R. Cardenas, ‘‘Distributed
[32] T. L. Vandoorn, J. De Kooning, B. Meersman, and L. Vandevelde, predictive control for frequency and voltage regulation in microgrids,’’
‘‘Voltage-based droop control of renewables to avoid on-off oscilla- IEEE Trans. Smart Grid, vol. 11, no. 2, pp. 1319–1329, Mar. 2020.
tions caused by overvoltages,’’ IEEE Trans. Power Del., vol. 28, no. 2, [54] B. Mirafzal, ‘‘Survey of fault-tolerance techniques for three-phase volt-
pp. 845–854, Apr. 2013. age source inverters,’’ IEEE Trans. Ind. Electron., vol. 61, no. 10,
[33] H. Zhang, S. Kim, Q. Sun, and J. Zhou, ‘‘Distributed adaptive virtual pp. 5192–5202, Oct. 2014.
impedance control for accurate reactive power sharing based on con- [55] C. Cecati, A. O. Di Tommaso, F. Genduso, R. Miceli, and G. Ricco
sensus control in microgrids,’’ IEEE Trans. Smart Grid, vol. 8, no. 4, Galluzzo, ‘‘Comprehensive modeling and experimental testing of fault
pp. 1749–1761, Jul. 2017. detection and management of a non-redundant fault-tolerant VSI,’’ IEEE
[34] J. Liu, Y. Miura, and T. Ise, ‘‘Comparison of dynamic characteristics Trans. Ind. Electron., vol. 62, no. 6, pp. 3945–3954, Jun. 2015.
between virtual synchronous generator and droop control in inverter-based [56] H. Akagi, ‘‘Classification, terminology, and application of the modular
distributed generators,’’ IEEE Trans. Power Electron., vol. 31, no. 5, multilevel cascade converter (MMCC),’’ IEEE Trans. Power Electron.,
pp. 3600–3611, May 2016. vol. 26, no. 11, pp. 3119–3130, Nov. 2011.

VOLUME 8, 2020 160535

 B. Mirafzal, A. Adib: On Grid-Interactive Smart Inverters: Features and Advancements

[57] Q. Yang, J. Qin, and M. Saeedifard, ‘‘A postfault strategy to control ASWAD ADIB (Member, IEEE) received the
the modular multilevel converter under submodule failure,’’ IEEE Trans. B.Sc. degree in electrical and electronic engineer-
Power Del., vol. 31, no. 6, pp. 2453–2463, Dec. 2016. ing from the Bangladesh University of Engineer-
[58] J. Lamb and B. Mirafzal, ‘‘Open-circuit IGBT fault detection and location ing and Technology (BUET), in 2012, and the
isolation for cascaded multilevel converters,’’ IEEE Trans. Ind. Electron., Ph.D. degree in electrical engineering from Kansas
vol. 64, no. 6, pp. 4846–4856, Jun. 2017. State University, Manhattan, KS, USA, in 2018.
[59] P. W. Hammond, ‘‘Enhancing the reliability of modular medium-voltage He worked as a Software Engineer with
drives,’’ IEEE Trans. Ind. Electron., vol. 49, no. 5, pp. 948–954, Oct. 2002.
Samsung Research and Development Institute
[60] P. Lezana and G. Ortiz, ‘‘Extended operation of cascade multicell con-
Bangladesh, from 2012 to 2014. In 2019,
verters under fault condition,’’ IEEE Trans. Ind. Electron., vol. 56, no. 7,
pp. 2697–2703, Jul. 2009. he worked as a Postdoctoral Fellow with Kansas
[61] J. Lamb and B. Mirafzal, ‘‘An adaptive SPWM technique for cascaded State University. He currently works on grid-interactive smart inverters as a
multilevel converters with time-variant DC sources,’’ IEEE Trans. Ind. Postdoctoral Researcher at Oak Ridge National Laboratory, TN, USA. His
Appl., vol. 52, no. 5, pp. 4146–4155, Sep. 2016. research interests include power electronics and advanced control schemes
for smart inverters.

BEHROOZ MIRAFZAL (Senior Member, IEEE)

received the Ph.D. degree in electrical engineering
from Marquette University, Milwaukee, WI, USA,
in 2005.
From 2005 to 2008, he was with Rockwell
Automation/Allen-Bradley, Mequon, WI, USA,
as a Senior Development/Project Engineer, where
he was involved in research and development
related to motor-drive systems. From 2008 to
2011, he was an Assistant Professor with Florida
International University, Miami, FL, USA. He is currently an Associate
Professor at Kansas State University, Manhattan, KS, USA. He has published
over 100 articles in professional journals and conferences and holds four
U.S. patents. His current research interests include the applications of power
electronics in grid-interactive converters and energy conversion systems. He
was a recipient of the 2008 Second Best IEEE Industry Applications Society
Transactions Prize Paper Award in 2007, the Best 2012 IEEE Power and
Energy Society Transactions Prize Paper Award in 2011, the 2014 U.S.
National Science Foundation CAREER Award, and the 2019 Frankenhoff
Outstanding Research Award, Kansas State University. He has served as
the Technical Co-Chair for the IEEE IEMDC Conference in 2009. He has
been serving as an Associate Editor for the IEEE TRANSACTIONS ON INDUSTRY
APPLICATIONS since 2011 and the IEEE TRANSACTIONS ON POWER ELECTRONICS
since 2018.

160536 VOLUME 8, 2020

View publication stats