Received 12 August 2022, accepted 16 September 2022, date of publication 26 September 2022, date of current version 4 October 2022.

Digital Object Identifier 10.1109/ACCESS.2022.3209650

An Adaptive Emergency Approach for Hybrid

Networked Microgrids Resilience
AYDA SHAKER 1,2 , (Student Member, IEEE), MOKHTAR BOZORG 2 , (Member, IEEE),
AMIN SAFARI 1 , AND SAJAD NAJAFI-RAVADANEGH 1 , (Member, IEEE)
1 Departmentof Electrical Engineering, Azarbaijan Shahid Madani University, Tabriz 53714-161, Iran
2 School
of Management and Engineering Vaud (HEIG-VD), University of Applied Sciences and Arts of Western Switzerland (HES-SO), 2800
Yverdon-les-Bains, Switzerland
Corresponding author: Ayda Shaker ([email protected])

ABSTRACT The low inertia of renewable-based distributed energy resources (DERs) renders hybrid
networked microgrids (NµGs) dynamically susceptible to transients. Such fragility makes it very difficult
for NµGs operators to maintain a reasonable margin for the resilient operation during extreme condition
contingencies. This paper presents a three-stage emergency approach to improve resilience of NµGs through
maintaining dynamic security. The proposed approach targets preserving the resilient operation of NµGs by
preventing unnecessary tripping of the DERs after unintentional islanding incident. To do so, a resilient
operation zone (ROZ) is introduced which determines the secure operating zone for NµGs and the limits
for implementing the corrective countermeasures for resilience augmentation. The proposed approach is
outlined in three stages: First, offline analysis is carried out to model and calculate the ROZ. Second, hybrid
NµGs operating point is monitored at the pre-event stage and the calculated ROZ at offline stage is adapted
to the operating conditions. The third stage is responsible for real-time evaluation of hybrid NµGs security
using the ROZ and implementation of the countermeasures. Comprehensive simulation studies presented in
this paper demonstrate effectiveness of the proposed scheme for enhancing resilience of hybrid NµGs.

INDEX TERMS Hybrid AC/DC microgrids, resilient power systems, emergency approach.

I. INTRODUCTION entire power system, the power system resilience could be

A. BACKGROUND further improved by networking the adjacent individual µGs
In recent years, extreme weather-related and man-made and establishing networked µGs (NµGs) [5]. A NµG might
events have led to more intense and frequent interruptions in include only AC or DC µGs; however, AC and DC µGs can
normal operation of conventional power systems [1]. Tremen- be practically networked to configure hybrid AC/DC NµGs
dous socio-economic consequences of such extreme events which can contribute more to the power system resilience.
have promoted the concept of power system resilience [2].
To establish a resilient power system, the conventional notion
B. LITERATURE REVIEW
of bulk power system utilization has been revisited as the
concept of distributed operation, say microgrids (µGs) [3]. During extreme condition contingencies, when the upstream
Deployment of grid-connected AC µGs with specific power supply is interrupted [6], the NµGs are exposed to
objectives, say hospitals, refineries, and alike, is a common an unintentional islanding incident which can challenge the
practice for power system resilience. By proliferation of resilient operation of the NµGs. The main reason is the low
renewable-based distributed energy resources (DERs) and physical inertia of power-electronic converters which render
DC loads, DC µGs are also became attractive [4]. Although hybrid NµGs dynamically fragile against incipient transients
individual AC and DC µGs may enhance the resilience of the and makes it very difficult for NµGs operators to maintain
a reasonable margin for the resilient operation [7]. Due to
The associate editor coordinating the review of this manuscript and such a vulnerability, certain procedures and grid codes have
approving it for publication was Maurizio Casoni . recommended fast DER tripping through the loss of mains

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
103164 VOLUME 10, 2022
A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

protection (ANSI Code 78) upon an unintentional island- C. RESEARCH GAPS AND CONTRIBUTIONS
ing event [8], [9]. Although DER tripping during extreme Based on the literature review study presented in Section I.A,
events might rescue the low-inertia DERs from likely dam- it can be observed that resilience enhancement through pre-
ages, it would significantly degrade the resilient operation serving dynamic security is worthy of study which is not cov-
of hybrid NµGs. The degraded NµGs and tripped DERs ered in the available literature, yet. Such an approach deems
might be restored through black-start mechanisms; however, vital since in case of dynamic insecurity and losing DERs,
this might be a very difficult task, particularly under severe black starting under severe weather conditions would be very
weather conditions [10]. To cope with this issue and to difficult. To fill this gap, an adaptive emergency approach is
establish a resilient operation: 1) The immediate tripping of presented in this paper which aims at expediting the resilient
DERs should be suspended; 2) The DERs should remain in operation of hybrid NµGs. The main contributions of this
service to the boundaries of dynamic insecurity [11]; and, paper are:
3) Proper emergency countermeasures should be adopted to
preserve the dynamic security, and in consequence, promote • Developing an analytical method, which is based on
region of attraction concept in non-linear control theory,
the resilient operation of the NµGs. In this regard, out-of-
to identify the dynamic security boundaries of a hybrid
step protection schemes are offered to distinguish the borders
NµGs.
of dynamic security [12]. Hence, the DERs can remain in-
• Proposing a suite of emergency approaches to maintain
service subsequent to an unintentional islanding event until
the DERs in-service before reaching the boundaries of
the trip signal is issued by out-of-step protection. However,
out-of-step relays are mostly applied to bulk power sys- dynamic insecurity. The proposed emergency counter-
measures adapt to different operation conditions of the
tems with considerable inertia and are deemed inefficient for
hybrid NµGs.
low-inertia µGs [13]. In [14], undervoltage relay is offered
• Embedding hazard characteristic in countermeasures
to detect the dynamic security boundaries. However, the
identification process;
lack of selectivity in voltage-based protection schemes can
also lead to unnecessary tripping of DERs. A new scheme • Exploiting the potential of NµGs facilities to prevent
based on the operation concept of existing overcurrent and the need for black-start and mitigate impacts of extreme
events.
undervoltage relays is offered in [15] to detect the borders
of dynamic security. The method proposed in [15] consid-
ers the normal contingencies such as short-circuit events II. PROPOSED METHODOLOGY
rather than extreme condition contingences, say unintentional A. OVERVIEW
islanding. The chronological outline of the proposed adaptive emer-
gency approach is depicted in Fig. 2. The main objective
is to preserve the resilient operation of NµGs by prevent-
ing unnecessary tripping of the DERs after an unintentional
islanding incident. To do so, a resilient operation zone (ROZ)
is introduced and the operation trajectories of the NµGs
are preserved within the ROZ through suite of emergency
approaches. The ROZ is the locus of all points within state
variables plan at which, the NµGs resilient operation is
retrieved subsequent to an unintentional islanding incident.
In the proposed method, unlike the conventional loss of mains
FIGURE 1. Performance of a power system subsequent to an extreme protection schemes, DERs can remain in-service and enhance
event.
resilience of the hybrid NµGs as long as the DERs state vari-
Most of the available researches in the literature for pre- ables lie within the boundaries of the ROZ. Only in case that
serving the dynamic security, and in consequence, boosting the boundaries of the ROZ are violated, the corresponding
up the resilience of NµGs, are control-oriented solutions [16], DER is tripped to avoid likely damages.
[17], [18], [19], [20], [21]. Although being effective to govern In Fig. 2, Stage 1 deals with offline analysis to calculate the
the dynamic security, such control approaches are usually ROZ. In this regard, the dynamic security model is extracted
triggered after dynamic security preservation stage. To offer first which describes the behavior of synchronous generator-
more details, the performance of a power system while facing based DERs (SGBDERs), inverter-based DERs (IBDERs),
an extreme event is depicted in Fig. 1. In Fig. 1, the emer- AC µGs, DC µGs, and hybrid NµGs while being subjected
gency dynamic security preservation stage is a very narrow to unintentional islanding incident. The attained model is
range within which, the emergency countermeasures should then used to calculate equilibrium points and the ROZ of the
be applied. However, the control mechanisms offered by [16], hybrid NµGs.
[17], [18], [19], [20], and [21] occurs afterwards, at the event The ROZ calculated at Stage 1 is dependent on the oper-
progress and degraded stages, when the dynamic insecurity ating point (loading level) of the NµGs which is handled
might already has happened. at the Stage 2 of the proposed approach. At Stage 2, the

VOLUME 10, 2022 103165

 A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

FIGURE 3. Block diagram of a synchronverter.

b, µB Index and set of buses

s, µ, ac Symbol of SGBDER, microgrid, AC
microgrid
e, m Symbol of electrical and mechanical
quantities
FIGURE 2. Proposed adaptive emergency approach for hybrid NµGs D, H Damping factor (p.u.), and inertia
resilience.
constant (second)
NµGs equilibrium points and associated ROZ are updated in E Voltage amplitude (p.u.)
accordance with the operating point of hybrid NµGs. In other P Active power (p.u.)
words, Stage 2 adapts the ROZ, which will be used at Stage 3, Y (θ) Modulus (argument) of admittance
to the operation condition of hybrid NµGs. In addition, a suite matrix (p.u.)
of countermeasures, which will be used in the aftermath of ω0 Nominal angular velocity (Rad./sec.)
the disaster, are determined in this stage. The analysis at δ, ω Rotor angle (Rad.) and rotor angular
Stage 2 is performed with a sufficient lead time for assurance velocity (Rad./sec.),
purposes. After extreme event condition, the state variables of 1, • Difference and derivative operators
hybrid NµGs are monitored at Stage 3. The monitored values
are compared with boundaries of the ROZ updated at Stage 2. 2) IBDER DYNAMICS
Based on the performed comparison, the proper decisions are The IBDERs controlled by the droop mechanism can be
selected and implemented from the countermeasures deter- considered equivalent to a synchronverter [23], [24]. The syn-
mined at Stage 2. chronverter mimics the dynamics of a synchronous generator
through the block diagram depicted in Fig. 3 and modeled
B. MATHEMATICAL BACKGROUND by (3)-(5) [25]:
The terms and definitions utilized in this paper are dis-
cussed in this section. Afterwards, the proposed approach is Tie = 1.5 mf if Ioi cos(δi − ϕi ) (3)
presented in Sections II.C and II.D using these terms and Qi = 1.5 mf if Ioi sin(δi − ϕi ) (4)
definitions. Ei = δ̇i mf if sin δi (5)

1) SGBDER DYNAMICS where,

Unintentional islanding from the main grid can be seen as a i Symbol of IBDER
relatively large disturbance from the standpoint of small-scale Io (ϕ) Output current (angle) of IBDER (p.u.)
SGBDERs in AC µGs. Hence, the equation of motion can be mf if Voltage set point of IBDER (p.u.)
used to describe such condition [22], that is (in per-unit): Q, T Reactive power (p.u.), torque (N.m)
 •
µ,ac
 δs

 = ω0 . 1ωsµ,ac Through the synchronverter model, the motion equation
• can represent the dynamics of IBDERs while being subjected
1 ωsµ,ac = (Pm,µ,ac
s − Pe,µ,ac
s − Dµ,ac
s 1ωsµ,ac ) (1)
 to severe disturbances [7]:
× (2Hsµ,ac )−1


 •
µ,ac µ,ac
 δi = ω0 . 1ωi

where,

•
µ µ,ac µ,ac
1ωi ,ac = (Pi (6)
m,µ,ac e,µ,ac
− Pi − Di 1ωi )
Pe,µ,ac
s = (Esµ,ac )2 Yssµ,ac cos θssµ,ac 
 µ,ac −1
× (2Hi )

µ,ac µ,ac
Eb Esµ,ac Ybs
X
+
where,
b ∈ µB
e,µ,ac µ,ac µ,ac µ,ac µ,ac
b 6= s Pi = 1.5 Efi Ioi cos(δi − ϕi ) (7)
µ,ac µ,ac
× cos(δsµ,ac − δb − θbs ) (2) where, Ef is the voltage set point of IBDER.

103166 VOLUME 10, 2022

A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

Theorem 1 (Lyapunov’s indirect method) [27]: Let A be a

Jacobian matrix of (8) at x∗ :
∂f
A= (12)
∂x x=x∗
The x∗ is an asymptotically stable equilibrium point of
(8) if all the eigenvalues associated with A are located on
the left-half plane which is denoted as secure equilibrium
point (SEP), hereinafter. Likewise, x∗ is unstable equilibrium
point of (8) if A has eigenvalues on the right-half plane
which is referred to as unsecure equilibrium points (UEPs),
hereinafter.
FIGURE 4. Hybrid AC/DC NµGs.

C. STAGE 1: OFFLINE ANALYSIS

3) RESILIENT OPERATION ZONE (ROZ) Given the hybrid AC/DC NµGs depicted in Fig. 4, this
The ROZ is defined as the locus of all points at which, the section represents the dynamics of a hybrid NµGs as the
NµGs resilient operation is retrieved when the disturbances system denoted by (8).
imposed by the unintentional islanding incident are cleared. The hybrid NµGs in Fig. 4 can be represented by the
In other words, ROZ determines up to when it is secure to dynamics of associated center of inertia as:
maintain the DERs in-service and stablish resilient operation.  •
The properties of ROZ fit well with region of attraction  δ N µ = ω . 1ωN µ
CoI 0 CoI
•
concept in non-linear control theory [26] which is discussed  1ω N µ = (2H N µ )−1 Pm,N µ − Pe,N µ − DN µ 1ωN µ
h i
CoI CoI CoI CoI
in the following.
Consider a nonlinear system model as: (13)
• where,
x = f (x) (8)  
Nµ Nµ µ,ac µ,ac
X
where, x us vector of state variables, f is a differentiable δCoI = (HCoI )−1  HCoI δCoI + H dc δ dc  (14)
function from a domain R1 ⊆ Rn into Rn . µ∈AC
Definition 1: ψ (t ; x0 ) is the solution of (8), with an initial  
value of x(0) = x0 , evaluated at time t ≥ 0. ψ (t ; x0 ) is the Nµ Nµ
X µ,ac µ,ac
1ωCoI = (HCoI )−1  HCoI 1ωCoI + H 1ω
dc dc 
system trajectory which crosses x0 .
µ∈AC
Definition 2: A vector x∗ ∈ Rn is an equilibrium point
of (8) if f (x∗ ) = 0. Here, ∗ is superscript for equilibrium (15)
Nµ µ,ac
X
point. HCoI =H dc
+ HCoI ,
Definition 3: The x∗ associated with (8) is: µ∈AC
1) a stable equilibrium point if for any ε ≥ 0, there exist a Nµ µ,ac
X
dc
DCoI =D + DCoI (16)
σ so that: µ∈AC

x0 − x∗ ≤ σ ⇒ ψ(t ; x0 ) − x∗ ≤ ε ∀t ≥ 0; (9) Pm,N µ = P m,dc

+Pm, ac
, Pe,N µ = Pe,dc + Pe, ac (17)
where, CoI is the subscripts for buses and center-of-inertia,
where, ε, σ are small positive values and kk is the norm
AC is set of AC µGs, and dc and Nµ are symbols of DC
operator.
microgrids and networked microgrids, respectively. In (14)
2) an asymptotically stable equilibrium point if (9) and (10) µ,ac µ,ac
and (15), δCoI and 1ωCoI are rotor angle and rotor angular
are satisfied:
velocity associated with center of inertia of each AC µG
x0 − x∗ ≤ σ ⇒ lim ψ(t ; x0 ) = x∗ (10) calculated as:
t→∞
µ,ac
X Hjµ,ac δjµ,ac
3) an unstable equilibrium point if (9) does not hold. δCoI = µ,ac ,
HCoI
Definition 4: The region of attraction associated with an j∈DERµ
asymptotically stable equilibrium point is:
µ,ac
X Hjµ,ac 1ωjµ,ac
1ωCoI = µ,ac (18)
HCoI
n o
< = x ∈ Rn | lim ψ (t ; x0 ) = x∗ (11) j∈ DERµ
t→∞
where,
< is a set of points, such that any trajectory originating from
µ,ac µ,ac µ,ac µ,ac
X X
x0 ∈ < at time 0 will be attracted to the stable equilibrium HCoI = Hj , DCoI = Dj (19)
point. j∈DERµ j∈DERµ

VOLUME 10, 2022 103167

 A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

dc(1)∗
and DERµ is set of DERs within the µth µG. In (18), DERµ 1ωi =0 (25)
set includes both SGBDERs and IBDERs within the µth AC µ,ac(1)(2)∗
xCoI =
X µ,ac µ,ac µ,ac(1)(2)∗
(HCoI )−1 Hj xj
µ,ac µ,ac
µG. Here if jth DER is a SGBDER, δj and 1ωj follow
j∈DERµ
the dynamics expressed by (1); otherwise, (6) represent the
µ,ac µ,ac (26)
dynamics of δj and 1ωj . Note that the batteries are
usually connected to the AC systems through an inverter. In (25), the negative sign corresponds to xi
dc(1)∗
and positive
In case the inverter of the battery is operated as the grid dc(2)∗
sign stands for xi . In (26), the index j encompasses both
forming DERs within an AC µG, the dynamics can also be
SGBDERs and IBDERs with in an AC µG. The discriminated
represented by (6). In case the batteries are operation in grid
vectors of equilibrium points for SGBDERs and IBDERs are
following mode, they can contribute to the countermeasures
as (27) and (28), shown at the bottom of the next page.
by rapidly charging and discharging which is discussed in µ,ac(1)∗
In (27) and (28), the negative sign corresponds to xs
Table 2. µ,ac(1)∗ µ,ac(2)∗
and xi ; and, the positive is related to xs and
The DC µGs in Fig. 4 are connected to the AC bus through µ,ac(2)∗
an interlinking converter. In other words, the DC NµGs are xi . The calculated equilibrium points, (25), (27),
seen as a large IBDER from AC bus standpoint with the and (28) are evaluated by Theorem 1 to identify associated
rating equal to sum of DERs rating connected to the DC bus. security status. To do so, Jacobian matrix of (13) is formed
Hence, δ dc and 1ωdc in (14) and (15) follow the dynamics by (29), shown at the bottom of the next page, and associated
represented in (6). In (14)-(16), H dc and Ddc are: eigenvalues are calculated by solving (30), shown at the
bottom of the next page, where, λ is matrix of eigenvalues.
µ,dc Nµ Nµ
X X
H dc = Hi ,
.
In (30), the term DCoI 4HCoI is a positive value; hence,
µ∈DC i∈DERµ to have the left-half plane:
µ,dc
X X
dc
D = Di (20)
Nµ 2
 
µ∈DC i∈DERµ
∂Pe,N µ DCoI
> (31)
where, DC is set of DC µGs. In (17), Pm,N µ represents Nµ
∂δCoI
Nµ
8ω0 HCoI
the total input power to the DERs within a NµGs. This can N µ(1)(2)∗
x=xCoI
be mechanical power for SGBDERs and DC power to the
µ,ac∗
IBDERs. The electrical and mechanical (input) quantities The requirement in (31) is fulfilled when the δidc∗ , δs ,
µ,ac∗
in (17) are calculated as: and δi in (25), (27) and (28) are less than π/2 radians.
µ.dc
X X m,µ,dc In (25), ϕi is close to π/2 radians since the IBDERs are
Pm,dc = Pi ,
usually connected to the microgrid through a relatively large
µ∈DC i∈DERµ dc(1)∗ dc(2)∗
coupling inductor. Hence, δi ≤ π 2 and δi ≥ π 2.
 
m,µ,ac µ,ac∗
X X
m,ac
P = Pj (21) In (27), the δs is calculated based on (2) where, Ybs
µ∈AC j∈ DERµ elements are zero, except those representing the link between
X X µ,dc µ,dc µ,dc µ,dc µ.ac
P e,dc
= 1.5Efi Ioi cos(δi − ϕi ) SGBDERs and the main AC bus, stated as YMs in (27).
µ.ac µ.ac
µ∈DC i∈DERµ The phase angle θMs associated with YMs is close to π/2
(22) radians representing high X/R ratio of SGBDERs and step-
µ,ac(1)∗ µ,ac(2)∗
X X e,µ,ac up transformers. Hence, δs is less than and δs
Pe,ac = Pj (23)
is larger than π/2 radians. In (28), the condition is the same
µ∈AC j∈ DERµ µ,ac(1)∗
as (25) and δi is less than π/2 radians. Therefore, based
Once the dynamic model of the hybrid NµGs is devised in on the requirements stated by Theorem 1 in Section II.B.3,
the form of (8), the ROZ can be calculated using Definition 4 N µ (1)∗ N µ (2)∗
xCoI and xCoI are SEP and UEP, respectively.
in Section II.B.3, (11). The first step is to calculate the SEPs Once the SEP and UEP are attained for hybrid NµGs, the
and UEPs which is performed by applying Definition 2 in region of attraction concept is calculated based on Definition
Section II.B.3 to (13). The equilibrium points of the hybrid 4 in Section II.B.3. The region of attraction associated with
NµGs represented by (13) are: N µ (1)∗
xCoI is an open and invariant set which is limited by the
N µ (1)(2)∗ N µ (2)∗
xCoI limit cycles of UEP, xCoI . Here, the limit cycle is formed
N µ (2)∗
P P µ,dc dc(1)(2)∗ P µ,ac µ,ac(1)(2)∗ by calculating the system trajectory crossing xCoI [26].
Hi xi + HCoI xCoI
µ∈DC i∈DERµ µ∈AC Fig. 5 depicts the schematical representation of the secure and
= N µ (2)∗
Nµ unsecure trajectories along with the limit cycle of xCoI for
HCoI
a hybrid NµGs demonstrated by (13). Here, the inner region
(24) of the limit cycle is the secure zone. In Fig. 5, the operating
N µ (1)∗
where, point of NµGs under normal operation conditions is xCoI .
m,µ.dc By NµGs islanding incident, the operating point of the NµGs
dc(1),(2)∗ dc(1),(2)∗ µ.dc Pi N µ (1)∗
xi : δi = ϕi ∓ cos−1 µ.dc µ.dc
, moves from xCoI towards the boundaries of secure zone,
1.5 Efi Ioi i.e. the limit cycle in Fig. 5.

103168 VOLUME 10, 2022

A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

at post-islanding condition. Here, the hurricanes are taken

into the account and the wind speed at pre-islanding stage is
the main monitored factor. The pre-islanding data monitoring
is repeated periodically with updating rate of TUpdate seconds
to adapt the emergency countermeasures to any change in
NµGs operation condition (blocks #3 and #4). Note that,
TUpdate is directly dependent on the polling and updating rate
of the NµGs data acquisition system and can be adjusted
based on the characteristic of NµGs data acquisition system.
In block #5, a fragility analysis is performed which cor-
relates the measured wind speed with the fragility curves of
NµGs facilities using (32) [28]:
FIGURE 5. Secure and unsecure trajectories for hybrid NµGs.   0 2
WS
Z WS −0.5× SD−1 ln Mean
e
ProbFailure = √ (32)dWS 0
0 2πSD × WS 0
Here, ‘‘Start’’ points indicate the locus of hybrid NµGs where, Mean and SD are mean and standard deviation values,
operating point at the time of remedial actions actuation. WS is wind speed (m/s), and ProbFailure is the probability of
Referring to Fig. 5, if the remedial actions are actuated when failure. Here, the fragility of wind-based DERs and the main
the operating points is ‘within the secure zone limits, the interconnecting links (MIL) which connect the µGs to the
resilient operation of NµGs can be maintained (green Start rest of the NµGs are taken in to the account. The main reason
point); otherwise, the situation yields in insecurity and all is that the failure in DERs would increase the power deficit
DERs within the NµGs should be tripped (red Start point). originated from unintentional islanding and consequently, the
amount of required remedial actions. On the other hand,
D. STAGE 2: PRE-EVENT MONITORING & UPDATING failure in MIL would change the topology of NµGs and
The outline of Stage 2 is depicted in Fig. 6. First step at hence, the available facilities to stablish resilient operation
Stage 2 is to monitor operating point of the NµGs, since of the NµGs.
the equilibrium points, associated region of attraction and The calculated failure probability in (32) is used to update
consequently, ROZ is highly dependent on the operating the dynamic security model, (13), based on Table 1. The
point of the NµGs. The operational factors which are mon- updated model augments performance of the proposed emer-
itored at block #1 in Fig. 5 are NµGs operating mode gency approach by representing more realistic mimic of post-
(importing/exporting energy), power trade of the NµGs with islanding condition. In Table 1, the failure for an asset is
upstream grid, transacted power of each µG with the rest concluded in case the failure probability calculated in (32)
of NµGs, and the operating point of the DERs. In addition, is greater than a pre-defined value, say 70%. This value
hazard condition is monitored to estimate the status of NµGs should be set by the NµGs operator through stablishing a

µ,ac(1),(2)∗

 1ωs
 =0
µ,ac
2
µ,ac µ,ac
xµ,ac(1),(2)∗
s µ,ac(1),(2)∗ µ,ac Ps
m,µ,ac
− EsYss cos θss (27)
 δs
 = θMs ∓ cos−1 µ,ac µ,ac
Es VM YMs
m,µ,ac
µ,ac(1),(2)∗ µ,ac(1), (2)∗ µ,ac P µ,ac(1)∗
xi : δi = ϕi ∓ cos−1 i
µ,ac µ,ac , 1ωi =0 (28)
1.5 Efi Ioi

ω0
" #
0
A = − ∂Pe,N µ (2H N µ )−1 Nµ Nµ
−DCoI (2HCoI )−1 (29)
Nµ CoI
∂δCoI N µ (1)(2)∗
x=xCoI
Nµ
ω0 ∂Pe,N µ

DCoI
λ 2
λ


 + N µ
+ Nµ Nµ
=0
∂δ




 2H CoI 2H CoI CoI N µ (1)(2)∗
 x=xCoI
v
!2 (30)
Nµ Nµ
2ω0 ∂Pe,N µ
u
DCoI 1u DCoI


λ = − Nµ ±


 t
N µ
− Nµ Nµ
2 ∂δ


 4H CoI 2H HCoI CoI CoI N µ (1)(2)∗

x=xCoI

VOLUME 10, 2022 103169

 A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

TABLE 1. Lookup table for model updating.

Nµ
Here, δCoI can be computed by placing (36) in (13) and
Nµ N µ∗
FIGURE 6. Outline of Stage 2 for the proposed emergency approach. solving the differential equation with δCoI = δCoI as the
initial condition:
tradeoff between the resilience and amount of remedial action
Nµ N µ∗
 
actuation. δCoI = r1 r2−1 t + r2−1 e−r2 t − r2−1 + δCoI ∀t ≥ 0
The updated model in Block #6 is then used to calculate the
(37)
equilibrium points using (24), determine associated security
attribute, and calculate the region of attraction (blocks #7 to
#9, respectively). Based on the region of attraction is calcu- If the NµGs is importing power from the main grid at pre-
lated in Block #9 of Fig. 6, the ROZ is determined in Block event condition, r1 is a negative value. Referring to (36) and
Nµ Nµ
#10. One may propose the utilization of the entire limit cycle (37), for r1 ≤ 0, both 1ωCoI and δCoI are monotonically
Nµ N µ∗ Nµ N µ∗
as an index, in which the inside and the outside of limit cycle decreasing within 1ωCoI ≤ 1ωCoI and δCoI ≤ δCoI .
would be labeled as blocking and tripping zones, respectively. In case of exporting power to the main grid, u2 is a positive
Nµ Nµ
However, it could be demonstrated that portions of limit value which yields both 1ωCoI and δCoI to be monotonically
Nµ N µ∗ Nµ N µ∗
cycles, designated as the ROZ, is sufficient for developing the increasing within 1ωCoI ≥ 1ωCoI and δCoI ≥ δCoI .
proposed resilience-oriented security measure. During grid- Therefore, the two portions of the limit cycle are sufficient
connected operation of NµGs, the total electrical power, i.e. to evaluate the hybrid NµGs security which are depicted in
Pe,N µ in (13), is equal to the Pm,N µ which is the sum of the Fig. 7.
power setpoint of IBDERs and mechanical input power of Block #11 in Fig. 6, determines the suite of countermea-
SGBDERs. On the other hand, Pe,N µ represents the net load sure to be used after the unintentional islanding scenario
of NµGs, i.e. is unfolded. The objective of these countermeasures is to
Nµ e,N µ alleviate the consequences of the disturbance before the time
Pe,N µ (t) = PLoad − (1 − u (t − τ )) PTrans (33)
that trajectories pass the ROZ boundaries and tripping of all
where, Trans is subscript for transacted value and u (t − τ ) DERs has happen. Table 2 summarizes the countermeasures
expresses the step endured by Pe,N µ while being subjected at Stage 2.
to NµGs islanding incident. In Fig. 1, the Pe,N µ is roughly In case of exporting power to the main grid, the coun-
constant during the emergency dynamic security preservation termeasure with the priority is to charge the battery storage
stage; on the contrary, Pe,N µ follows the system dynamics. systems. In case of inadequacy, the next priority is prompt
For emergency dynamic security preservation stage in Fig. 1, cutting down (not tripping) the outputs of IBDERs (includ-
• ing DERs at DC side) [29]. This can reduce the generation
Nµ
1 ωCoI (13) can be expressed as: excess and maintain resilient operation without DER tripping.
•
If the amount of IBDER curtailment is not sufficient, some
Nµ Nµ SGBDERs might also be tripped as the second priority. The
1 ωCoI = r1 − r2 1ωCoI (34)
proposed method to determine sufficiency of a countermea-
where, sure for resilient operation is presented in Section II.E. In case
of importing power from the grid, the priority is to rapidly
Nµ Nµ Nµ
 
r1 = (2HCoI )−1 Pm,N µ − Pe,N µ , r2 = (2HCoI )−1 DCoI discharge the battery storages. In case of insufficiency, the
(35) next priority is rapid load shedding is used where the load
shedding priority will be defined by NµGs operator.
Solving the differential equation represented by (34), and This stage monitors the state variables of the NµGs, i.e.
Nµ N µ∗ Nµ Nµ
using 1ωCoI = 1ωCoI as the initial condition yields: δCoI and 1ωCoI , and maps the trajectory within the ROZ
determined at Stage 2. Once the trajectory approaches the
Nµ
1ωCoI = r1 r2−1 (1 − e−r2 t ) ∀t ≥ 0 (36) boundaries of the secure zone, the countermeasures defined

103170 VOLUME 10, 2022

A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

FIGURE 8. Steps of countermeasures actuation.

FIGURE 7. ROZ for hybrid NµGs operation.

TABLE 2. Lookup table for post-islanding condition estimation. III. SIMULATION RESULTS
This section examines the proposed scheme on a system
depicted in Fig. 10.
The system data are available in [4] where µGs1 and 2 are
considered as the AC and the rest are DC µGs. The DER
installed capacity and the peak load associated with each
are reported in Table 3. The studied cases are represented
in Table 4. Here, the α in (38) is considered 0.7 which is
determined based on the dynamics of the hybrid NµGs under
study and simulation studies. The updating rate of NµGs with
by Table 1 at Stage 2 will be realized. This is modeled as: data acquisition system is assumed to be 1 second. Hence,
Nµ TUpdate in Fig. 6 is considered to be 1 second. In this study,
1ωCoI ≤ α 1ωLC |δ Nµ (38) the simulations are conducted using the DIgSILENT Power
LC =δCoI

where, (δLC , 1ωLC ) represents an ordered pair corresponding Factory software in a personal computer with Intel CoreTM i7
to the limit cycle (boundaries of the secure zone). 0 ≤ α ≤ CPU @3 GHz and 12 GB RAM.
1 determines the time for the countermeasures actuation. The simulation results for Cases I and II in Table 4 are
Here, α = 1 represents the ROZ in Fig. 8 which corresponds presented in Figs. 11 and 12. At pre-islanding stage, the
to the theoretical threshold. By reducing α, the resilience NµGs is operation point on associated SEP; hence, the locus
N µ (1)∗
operation zone diminishes which allows us to consider a of NµGs state variables is on the xCoI in Fig. 11. Here, the
safety margin to compensate practical inaccuracies (e.g. time required to form ROZ at pre-islanding stage is 600 ms
delays in communications, measurement errors, etc.). With and 400 ms for Cases I and II, respectively. Following to
respect to the practical considerations, large value for α may an islanding incident, the state variables move toward the
Nµ N µ∗ Nµ
result in late actuation of the countermeasures which endan- boundaries of ROZ within 1ωCoI ≤ 1ωCoI and δCoI ≤
N µ∗ Nµ
gers the resilience of NµGs. On the contrary, small α val- δCoI region for Case I (Fig. 11(a)) and within 1ωCoI ≥
N µ∗ Nµ N µ∗
ues may increase the sensitivity of the proposed ROZ-based 1ωCoI and δCoI ≥ δCoI region for Case II (Fig. 11(b)). This
approach to the transients and the countermeasures might observation is in line with the discussion made in Fig. 7.
be unnecessarily actuated for small disturbances. Therefore, In Case I, the pre-islanding energy trade between NµGs
a trade-off by should be established in determining the α and the main grid is 5.2 MW (import). Referring to Fig. 11(a),
value in a way that neither the security nor the dependability in case the 5.2 MW load is curtailed at T1, which the time
of the proposed scheme will be jeopardized. In the proposed NµGs trajectories exceed the boundaries of countermeasures
approach, α is considered as the setting which should be set actuation limit, the NµGs trajectories are steered towards
by the NµGs operator\decision maker. the SEP and resilient operation of NµGs can be retrieved.
The required amount of countermeasures to be applied This can also be observed from temporal characteristic of
at the post-islanding stage is equal to the amount of NµGs rotor angular velocity at NµGs center of inertia in Fig. 12(a).
power transaction with the main grid at pre-islanding stage On the contrary, resilient operation of NµGs is forfeited when
(assuming that load conditions have not changed from pre- the 5.2 MW load curtailment is occurred at T2 in Fig. 11(b),
to post-event stages). The countermeasures are actuated by i.e. beyond the ROZ. Here, the pole slip event in Fig. 12(b)
violating the brown dashed line in Fig. 8. yields in insecurity which in turn, results in tripping of all
The load curtailment-based countermeasures can be DERs and losing 10 MW load.
directly applied. However, for DER curtailment-based coun- In Case II, 4.45 MW was exporting at pre-islanding stage
termeasure, the curtailment occurs through the logic depicted where corresponding countermeasure to maintain NµGs
in Fig. 9 which combines local decision with NµG-wide resilience is 4.45 MW DER curtailment. In Fig. 11(b), the
decision: resilient operation of the NµGs is preserved by curtailing

VOLUME 10, 2022 103171

 A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

FIGURE 9. Logic of DER curtailment (in case needed) subsequent to

unintentional islanding event.

FIGURE 11. Trajectories of NµGs for: a) Case I; b) Case II.

FIGURE 10. 33-bus distribution system composed of four interconnected

µGs operating as hybrid NµGs [4].

TABLE 3. DERs and peal load at each µG.

FIGURE 12. Temporal characteristic for rotor angular velocity at NµGs

center of inertia: a) Case I – load shedding at T1; b) Case I – load
shedding at T2; c) Case II – DER curtailment at T3; d) Case II – DER
curtailment at T4.
TABLE 4. Cases under study.

To validate the proposed method adaptability, the simu-

lation results for different cases are reported in Table 5 are
presented in Table 6. Here, the performance of the proposed
scheme (PS) is compared with out-of-step (OOS) relay-based
[13] and undervoltage (UV) relay-based [14] approaches as
4.45 MW DER curtailment at T3, where the NµGs trajec- the common loss of mains practices. The settings of OOS
tories violate the boundaries of countermeasures actuation relay is derived from [30], i.e. DER tripping after one pole slip
limit. Doing so, the resilient operation of NµGs is pro- incident. For UV relay, the settings are 0.8 p.u. with 200 ms
vided and 3.8 MW load is served at post-islanding stage. delay as recommended by [13]. Note that, the main objective
However, if 4.45 MW is curtailed outside the ROZ, i.e. of the proposed method is maintaining the resilience through
T4 in Fig. 11(b), the resultant situation yields in insecu- supply continuity. Hence, the main index used in Table 6 for
rity. The insecurity is emerged as the pole slip incident in comparison is the amount of load which curtailed/rescued
Fig. 12(d) which brings about tripping of all DERs and after unintentional islanding incident.
losing 3.8 MW load. As can be seen, by the proposed In Table 6, PS rescued considerable amount of load in all
ROZ-based approach in place, the resilient operation of the listed cases. In cases where the NµGs is importing power
NµGs subsequent to an unintentional event is maintained from the main grid, i.e. Cases I, III, IV, IX, and X, the imme-
at both Cases I and II. Referring to Fig. 12, both Cases I diate tripping of DERs through the local loss of mains relays
and II are settled down roughly 15 second after the island- are avoided and DERs are maintained in service through the
ing incident by actuating the remedial actions within ROZ logic depicted in Fig. 9. Doing so, the considerable amount of
limits; however, such settlements require emergency actions load, which is equal to the in-service on-site DERs, is rescued.
during dynamic security preservation stage (see Fig. 1). The In cases where the NµGs is exporting power to the main grid,
main reason is low inertia of hybrid NµGs which results in i.e. Cases II, V, and VI, 100% of the load is rescued upon
fast transients. Comparing Fig. 12(a) and (b), if the reme- unintentional islanding. The main reason is availability of
dial action in T1 is applied 150 ms later, that is T2, the sufficient on-site DERs which are kept in-service through the
resilient operation can not be maintained. Based on these proposed approach and contributed to resilient operation. The
observations, it can be concluded that the proposed ROZ is results outlined in Table 4 demonstrate that PS is capable of
an effective tool to discriminate the secure operating zone adapting to different operation condition of NµGs. Unlike the
of the NµGs and the limits for implementing the corrective PS, the deployment of OOS and UV schemes in most cases
countermeasures. resulted in substantial load curtailment.

103172 VOLUME 10, 2022

A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

TABLE 5. Cases under study.

TABLE 6. Simulation results for different cases.

FIGURE 15. Rescued load after unintentional islanding for different DER
generation and exchange with main grid conditions.

FIGURE 16. Rescued load after unintentional islanding for different

network structure.

Table 6 implies the undesirable impact of the available loss

of mains protection schemes on the resilience of the NµGs.
Case X in Table 5 expresses a scenario where unintentional
FIGURE 13. Temporal characteristic for voltage at main AC bus. islanding is occurred at extreme wind condition. For this case,
the wind speed is given 50 m/s. The fragility curves for wind-
based DERs and overhead lines connecting buses 6, 7 and
DC NµGs to the main AC bus are depicted in Fig. 14 [28].
Referring to Fig. 14, 50 m/s wind speed results in failure of
DER1 in AC µG1 and also splitting of NµGs from each other
due to failure of main interconnecting links. Here unlike other
cases, the load curtailment by the PS is 5.5 MW which is
more than NµGs power transaction with the main grid at pre-
islanding stage, 4.7 MW. The main reason is that according
FIGURE 14. Fragility curves for wind turbines and overhead lines. to Table 1, the generation of DER1 (0.8 MW) is added on
top of the NµGs power transaction with the main grid at pre-
islanding stage.
In Case VII, where no active power and low reactive power, Fig. 15 depicts the rescued load after unintentional island-
i.e. 100 kVAr, exchange with the main grid is envisioned, ing for different DER generation and exchange with main grid
deployment of OOS and UV schemes contributed to no load conditions. In Fig. 15, negative values for exchange represent
curtailment. However, when high reactive power exchange is the export to the main grid where, the resilient operation of
required (6 MVAr), 100% of the load is curtailed by UV. The the NµGs is fully attained and all loads within the NµGs is
main reason is the drastic voltage drop which occurs after rescued.
unintentional islanding at Case VIII (see Fig. 13). In Fig. 13, The amount of rescued load decreases as the DER gener-
the voltage amplitude is less than 0.8 p.u. for more than ation decreases which confirm the direct impact of on-site
200 ms which renders UV scheme to trip the DERs. However, DERs on the resilience of power system. When the exchange
even under such a condition, the PS can rescue the loads and with the main grid turns into positive values, the amount
boost up resilience of the NµGs. The simulation results in of rescued load decreases. The reason is that the positive

VOLUME 10, 2022 103173

 A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

exchange values account for import from the main grid mode the uncertainties originated from probability-based decision
that is equivalent to excess of load from the on-site DER making may reduce the precision level of decision making for
generation. This observation reveals the necessity of adopting the countermeasures.
an adaptive approach, such as PS, to maintain the resilient Future works may consider the scalability issues regarding
operation of NµGs. the type and the number of components of the NµGs and
Fig. 16 represent the performance of the PS in Case I communication standpoints. In particular, the effect of elec-
of Table 6 for different network configurations. Here, five trical vehicle, for instance vehicle to grid model, could be
configurations are considered for NµGs depicted in Fig. 10. investigated. Furthermore, the performance of the proposed
As can be seen, the PS can improve power system resilience approach in real-world applications could be studied through
even if the µGs are not networked. However in Fig. 16, experimental investigation. To this end, the authors aimed
the amount of load curtailment is reduced as we increased at performing experimental validation tests in the reconfig-
the connectivity among the µGs which in turn, increases the urable distribution grid laboratory of HEIG-VD, in Yverdon-
transacted power among the µGs. This observation expresses les-Bains, Switzerland [31].
effectiveness of the NµGs for augmenting power system
resilience and reducing the damaging impacts of electricity REFERENCES
interruptions. [1] X. Liu, M. Shahidehpour, Z. Li, X. Liu, Y. Cao, and Z. Bie, ‘‘Microgrids for
enhancing the power grid resilience in extreme conditions,’’ IEEE Trans.
Smart Grid, vol. 8, no. 2, pp. 589–597, Mar. 2017.
IV. CONCLUSION
[2] S. Teimourzadeh, O. B. Tor, M. E. Cebeci, A. Bara, and S. V. Oprea,
This paper dealt with augmentation of NµGs resilience ‘‘A three-stage approach for resilience-constrained scheduling of net-
through maintaining dynamic security. Here, a three-stage worked microgrids,’’ J. Modern Power Syst. Clean Energy, vol. 7, no. 4,
emergency approach was proposed which aims at preserving pp. 705–715, Jul. 2019.
[3] Z. Li, M. Shahidehpour, F. Aminifar, A. Alabdulwahab, and Y. Al-Turki,
the resilient operation of NµGs by preventing unnecessary ‘‘Networked microgrids for enhancing the power system resilience,’’ Proc.
tripping of the DERs after unintentional islanding incident. IEEE, vol. 105, no. 7, pp. 1289–1310, Jul. 2017.
A resilient operation zone (ROZ) was introduced and the [4] L. Che and M. Shahidehpour, ‘‘DC microgrids: Economic operation and
enhancement of resilience by hierarchical control,’’ IEEE Trans. Smart
operation trajectories of the NµGs are preserved within the Grid, vol. 5, no. 5, pp. 2517–2526, Sep. 2014.
ROZ through suite of emergency approaches. The conducted [5] A. Shaker, A. Safari, and M. Shahidehpour, ‘‘Reactive power management
studies in this paper concluded that: 1) The available loss of for networked microgrid resilience in extreme conditions,’’ IEEE Trans.
Smart Grid, vol. 12, no. 5, pp. 3940–3953, Sep. 2021.
mains protection schemes could jeopardize the resilient oper-
[6] Z. Ye, C. Chen, B. Chen, and K. Wu, ‘‘Resilient service restoration
ation of the NµGs; 2) By adopting suite of proper counter- for unbalanced distribution systems with distributed energy resources by
measures, such as the proposed approach, the negative impact leveraging mobile generators,’’ IEEE Trans. Ind. Informat., vol. 17, no. 2,
of loss of mains protection schemes on resilient operation pp. 1386–1396, Feb. 2020.
[7] S. Teimourzadeh, F. Aminifar, M. Davarpanah, and M. Shahidehpour,
of NµGs can be compensated for; 3) The proposed ROZ is ‘‘Adaptive protection for preserving microgrid security,’’ IEEE Trans.
an effective tool to discriminate the secure operating zone Smart Grid, vol. 10, no. 1, pp. 592–600, Jan. 2019.
of the NµGs and the limits for implementing the corrective [8] IEEE Standard for Interconnecting Distributed Resources With Electric
Power Systems, IEEE Standard 1547-2003, 2003, pp. 1–28.
countermeasures for resilience augmentation; 4) The amount [9] F. Iov, A. D. Hansen, P. E. Sorensen, and N. A. Cutululis, ‘‘Mapping of grid
of rescued load subsequent to an unintentional islanding event faults and grid codes,’’ Riso Nat. Lab., Tech. Univ. Denmark, Roskilde,
is highly dependent on the operation point of NµGs at pre- Denmark, Tech. Rep. Risø-R-1617, 2007.
[10] L. Zhang, C. Wang, J. Liang, M. Wu, B. Zhang, and W. Tang,
islanding stage which is tackled through adaptive feature
‘‘A coordinated restoration method of hybrid AC/DC distribution net-
of the proposed approach; 5) The proposed approach can work for resilience enhancement,’’ IEEE Trans. Smart Grid, early access,
improve the resilience even if the µGs are not networked; Jul. 21, 2022, doi: 10.1109/TSG.2022.3192910.
however, networking µGs and stablishing NµGs adds to the [11] S. Teimourzadeh, F. Aminifar, M. Davarpanah, and M. Shahidehpour,
‘‘Adaptive control of microgrid security,’’ IEEE Trans. Smart Grid, vol. 9,
effectiveness of the proposed approach in further improve- no. 4, pp. 3909–3910, Jul. 2018.
ment of power system resilience. [12] K. Malmedal, P. K. Sen, and J. P. Nelson, ‘‘Application of out-of-step
Regarding the limitations of the proposed approach, it is relaying for small generators in distributed generation,’’ IEEE Trans. Ind.
Appl., vol. 41, no. 6, pp. 1506–1514, Nov. 2005.
worth mentioning that the proposed approach relies on [13] S. Teimourzadeh, F. Aminifar, and M. Davarpanah, ‘‘Microgrid dynamic
the estimation of ROZ. The precision of estimated ROZ security: Challenges, solutions and key considerations,’’ Electr. J., vol. 30,
may depend on the accuracy of NµGs dynamic security no. 4, pp. 43–51, May 2017.
[14] E. J. Coster, J. M. A. Myrzik, and W. L. Kling, ‘‘Influence of protection
model (13). For the presented NµGs it is shown that the on transient stability of medium voltage grids including distributed gen-
approximated dynamic security model is sufficiently accu- eration,’’ in Proc. 42nd Int. Universities Power Eng. Conf., Sep. 2007,
rate (see figure 11). In case of application of the proposed pp. 1054–1059.
method to a more complex NµGs, accuracy of the dynamic [15] R. Razzaghi, M. Davarpanah, and M. Sanaye-Pasand, ‘‘A novel protective
scheme to protect small-scale synchronous generators against transient
security model of the NµGs must be carefully verified. instability,’’ IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1659–1667,
The other limitation might be deployment of fragility curve Apr. 2013.
for estimating the post-islanding condition which imposes [16] Y. Zhang, Q. Sun, J. Zhou, L. Li, P. Wang, and J. M. Guerrero, ‘‘Coor-
dinated control of networked AC/DC microgrids with adaptive virtual
probability-based decision making for post-islanding con- inertia and governor-gain for stability enhancement,’’ IEEE Trans. Energy
ditions. Although the proposed approach is still applicable, Convers., vol. 36, no. 1, pp. 95–110, Mar. 2021.

103174 VOLUME 10, 2022

A. Shaker et al.: Adaptive Emergency Approach for Hybrid Networked Microgrids Resilience

[17] Q. Hui, J. Yang, X. Yang, Z. Chen, Y. Li, and Y. Teng, ‘‘A robust control MOKHTAR BOZORG (Member, IEEE) received the B.Sc. and M.Sc.
strategy to improve transient stability for AC–DC interconnected power degrees in electrical engineering from the Sharif University of Technology,
system with wind farms,’’ CSEE J. Power Energy Syst., vol. 5, no. 2, Tehran, Iran, in 2008 and 2011, respectively, and the Ph.D. degree from the
pp. 259–265, Jun. 2019. Power System Research Group, Ecole Polytechnique Fédérale de Lausanne
[18] X. Li, Z. Li, L. Guo, J. Zhu, Y. Wang, and C. Wang, ‘‘Enhanced dynamic (EPFL), Lausanne, Switzerland, in 2015. From 2015 to 2018, he was a
stability control for low-inertia hybrid AC/DC microgrid with distributed Postdoctoral Fellow and the Guest Scientist at the Distributed Electrical
energy storage systems,’’ IEEE Access, vol. 7, pp. 91234–91242, 2019. System Laboratory, EPFL. Since 2019, he has been an Associate Professor of
[19] Q. Zhou, M. Shahidehpour, A. Alabdulwahab, and A. Abusorrah, ‘‘Flex-
energy and power systems with the School of Management and Engineering
ible division and unification control strategies for resilience enhance-
(HEIG-VD), Institute of Energy and Electrical Systems (IESE), University of
ment in networked microgrids,’’ IEEE Trans. Power Syst., vol. 35, no. 1,
pp. 474–486, Jan. 2020. Applied Sciences and Arts of Western Switzerland (HES-SO). His research
[20] L. He, Y. Li, J. M. Guerrero, and Y. Cao, ‘‘A comprehensive inertial control interests include smart grids, active distribution networks, applications of
strategy for hybrid AC/DC microgrid with distributed generations,’’ IEEE mathematical modeling, optimization techniques, data analytics in power
Trans. Smart Grid, vol. 11, no. 2, pp. 1737–1747, Mar. 2020. systems, integration of renewable energy sources, and energy storage sys-
[21] Y. Du, X. Lu, B. Chen, and F. Lin, ‘‘Resiliency augmented hybrid AC and tems into power systems.
DC distribution systems with inverter-dominated dynamic microgrids,’’
IEEE Trans. Smart Grid, vol. 13, no. 5, pp. 4088–4101, Sep. 2022.
[22] P. Kundur, N. J. Balu, and M. G. Lauby, Power System Stability and
Control. New York, NY, USA: McGraw-Hill, 1994.
[23] J. Rocabert, A. Luna, F. Blaabjerg, and P. Rodriguez, ‘‘Control of power
converters in AC microgrids,’’ IEEE Trans. Power Electron., vol. 27,
no. 11, pp. 4734–4749, Nov. 2012.
[24] S. D’Arco and J. A. Suul, ‘‘Equivalence of virtual synchronous machines
and frequency-droops for converter-based microgrids,’’ IEEE Trans. Smart
Grid, vol. 5, no. 1, pp. 394–395, Jan. 2014.
[25] Q.-C. Zhong and G. Weiss, ‘‘Synchronverters: Inverters that mimic
synchronous generators,’’ IEEE Trans. Ind. Electron., vol. 58, no. 4,
pp. 1259–1267, Apr. 2011.
[26] H. Khalil, Nonlinear Systems. New York, NJ, USA: Prentice-Hall, 2002. AMIN SAFARI received the B.Sc. degree in electrical engineering from the
[27] S. H. Zak, Systems and Control. New York, NY, USA: Oxford Univ. Press, University of Tabriz, Tabriz, Iran, in 2007, the M.Sc. degree from the Uni-
2003. versity of Zanjan, in 2009, and the Ph.D. degree from the Iran University of
[28] R. Mo, H. Kang, M. Li, and X. Zhao, ‘‘Seismic fragility analysis of
Science and Technology, Tehran, Iran, in 2013. He is currently an Associate
monopile offshore wind turbines under different operational conditions,’’
Professor with the Department of Electrical Engineering, Azarbaijan Shahid
Energies, vol. 10, no. 7, p. 1037, Jul. 2017.
[29] Y. Huang, G. Li, C. Chen, Y. Bian, T. Qian, and Z. Bie, ‘‘Resilient Madani University, Tabriz. His research interests include the application
distribution networks by microgrid formation using deep reinforcement of artificial intelligence, heuristic optimization algorithms to power system
learning,’’ IEEE Trans. Smart Grid, early access, Jun. 1, 2022, doi: design, FACTS, power system analysis and control, renewable energy, con-
10.1109/TSG.2022.3179593. trol and management of microgrids, and smart grid. In 2016, he was selected
[30] D. Reimert, Protective Relaying for Power Generation Systems. as a Distinguished Researcher at the Azarbaijan Shahid Madani University.
Boca Raton, FL, USA: CRC Press, 2005.
[31] M. Carpita, J.-F. Affolter, M. Bozorg, D. Houmard, and S. Wasterlain,
‘‘ReIne, a flexible laboratory for emulating and testing the distribution
grid,’’ in Proc. 21st Eur. Conf. Power Electron. Appl. (EPE ECCE Eur.),
Sep. 2019, pp. 1–6.

AYDA SHAKER (Student Member, IEEE) received the B.Sc. degree in

electrical engineering and the M.Sc. degree (Hons.) in electrical engineering
from the University of Tabriz, Iran, in 2016 and 2018, respectively. She
is currently pursuing the Ph.D. degree with the School of Engineering,
Azarbaijan Shahid Madani University, Tabriz, Iran. Since 2020, she has
been collaborating with EPRA Electric Energy Company, Ankara, Turkey,
as a Research and Development Researcher. She is also affiliated with SAJAD NAJAFI-RAVADANEGH (Member, IEEE) received the Ph.D.
the School of Management and Engineering Vaud (HEIG-VD), Institute of degree in electrical engineering from the Department of Electrical Engi-
Energy and Electrical Systems (IESE), University of Applied Sciences and neering, Amirkabir University of Technology (Tehran Polytechnic), Tehran,
Arts of Western Switzerland (HES-SO), Yverdon-les-Bains, Switzerland, Iran, in 2009. He is currently a Professor with the Electrical Engineering
as an Exchange Ph.D. Research Scholar. Her research interests include power Department, Azarbaijan Shahid Madani University, Tabriz, Iran, where he is
system operation, adaptive protection, microgrid resilience, and smart grid responsible for the Resilient Smart Grids Research Laboratory.
initiatives.

VOLUME 10, 2022 103175