This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
1

Microgrid: Configurations, Control and Applications

M. Rezkallah, Member, IEEE, A. Chandra, Fellow, IEEE, B. Singh, Fellow, IEEE, and S. Singh, Senior Member, IEEE

 [1]. Compared to DG, WT or MHP driven squirrel-cage

Abstract—This paper deals with a comprehensive study of induction machine (SCIM) with a capacitor bank [5-6] are
configurations of microgrid (MG) based on variable speed sustainable, and they require additional control for fixed
generators. The selection criteria for these configurations along frequency and voltage at PCC. To ensure continuous power to
with the design of required control algorithms, are presented in the load, hybrid of two or more ESs is suggested in [7-10]. Of
detail. Simulation results and experimental validation of selected course, this solution is effective instead of using only one ES
MG configurations are given to demonstrate their suitability. It is
aimed to enhance applications of MGs with main emphasis on
in MG. The drawbacks related to continuous operation of
renewable energy sources to ensure sustainable development in fixed speed DGs with varying loads, as well as, efficiency
rural and remote areas. maximization of RESs energy conversion still exist. The
performance obtained in [10] using WT and DG with variable
Index Terms—Microgrid (MG), dispersed generation units speed operation show considerable improvement in terms of
(DGUs), variable speed generators (VSGs), control design. reduced fuel consumption of DG and increased WT efficiency.
There are many such combinations possible involving various
I. INTRODUCTION RESs for continuous supply to connected loads in a MG.

H YBRID microgrids (MGs) are off-the-grid systems.

Electrical energy in such MGs, is generated using
available renewable energy sources (RESs) such as solar
Similar solutions having VSGs are presented here.
Generally, MGs use AC bus to tie all available ESs and
utilize the existing AC power system standards for the design,
photovoltaic (PV) array, wind turbine (WT), micro-hydro control, and protection [9]. Other option is the use of DC bus
power (MHP), geothermal source (GS), thermoelectric as it has advantages of higher reliability, simple control, and
generator (TG), tidal or wave energy generator (TWG), efficient interfacing with DC RESs, DC loads, and BESS [11].
biomass energy generators (BEG) and a diesel generator (DG) To capitalize advantages of both type of buses, hybrid
as a conventional back up energy source (ES). MGs comprise configuration is recommended in [12-13]. Using hybrid
of power converters, battery energy storage system (BESS), configuration, it becomes easy to connect various ESs driven
control and protection systems. Amongst various RESs, PV is generators at DC bus and use only one interfacing VSC to tie
all ESs to AC bus at PCC [14-18]. MG is a self-sustainable
generally tied to the AC bus known as point of common
system having various ESs for electricity generation. Such
coupling (PCC) through VSC (Voltage Source Converter),
installation serves electricity requirements for data centers,
whereas, other RESs use generators for energy conversion,
hospitals, industries, communication and petroleum stations,
which are interfaced to PCC either directly or through power military bases, isolated villages, mine sites, etc. Its size
converters. Variable speed generators (VSGs) have advantages depends on connected loads, whereas, output voltage and
as compared to fixed speed generators from efficiency point of current determine categories of its configuration, either DC or
view, however, at increased control complexity and cost [1-3]. AC. Recent research [19-20] is associated with control and
Despite breakthrough developments in power converters and power management strategies of the MG, whereas, consistent
control, fixed speed DG using synchronous machine (SM) and focused approach is required on design of MGs, where the
remains principle ES in isolated and remote areas [4]. The DG cost of installation, use of power converters and integration of
energy is costly and highly polluting, contributing to CO2 special machines, as generators are prime considerations.
emissions and global warming. In addition, the fuel is To emphasize these aspects, a comprehensive study of MGs
extremely expensive in remote areas due to the cost of based on VSG technology is presented here. These MGs are
transportation and safe storage requirements. Diesel engines divided into 11 broad categories having 6 configurations each,
(DEs) are noisy and suffer from premature wear when totalling to 66 MGs with associated control and special
operated at light loads, thereby increased maintenance cost machines as generators for different RESs. The major
contributions of this work are (1) proposed configurations
have only one VSC at PCC; (2) reduction in control
complexity and cost of the system; (3) maintenance free
operation because of many brushless generator configurations;
M.Rezkallah, A. Chandra, and S. Singh are with Department of Electrical
Engineering, École de Technologie Supérieure, 1100 Notre-Dame Montréal,
(4) application potential in diverse operating and weather
Québec H3C1K3 Canada. (miloud.rezkallah.1@ens.etsmtl.ca; conditions. These configurations are presented to provide an
ambrish.chandra@etsmtl.ca, sanjeev.chauhan@etsmtl.ca ). exposure to new control strategies and application potentials
B. Singh is with Department of Electrical Engineering, Indian Institute of of RESs in rural and isolated areas for sustainable growth.
Technology Delhi, New Delhi-110016, India. (bsingh@ee.iitd.ac.in).

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
2

II. MG CONFIGURATIONS
These MGs are classified on a basis of sources used as two,
three, and four energy sources. These MGs are selected on the
basis of power, available energy sources in remote areas and
the machine used as a generator for DGUs.
In these MGs, most DGUs are tied to a DC bus using
suitable converters except the DGUs having DFIG (Doubly
Fed Induction Generator), to avoid synchronization issues
with AC bus and to minimize the number of power converters.
The DFIG stator is directly connected to the AC bus whereas
its rotor is connected to DC bus through VSC, to extract
power from the rotor. Amongst other DGUs, the PV is
connected to DC bus using a DC-DC converter, whereas, SG
(Synchronous Generator), SCIG (Squirrel Cage Induction
Generator), SyRG (Synchronous Reluctance Generator),
Fig. 1 MG configurations based on PV and WT driven: a) SG, b) SyRG, c)
PMSG (Permanent magnet Synchronous Generator), and PMBLDCG, d) SCIG, e)PMSG, and f)DFIG
PMBLDCG (Permanent magnet brushless DC Generator) are
2) Configurations Based on PV and DG
connected to DC bus through VSC. The PMSG and
Fig.2 shows MGs based on PV and DE driven VSG
PMBLDCG are also connected to DC bus through a diode
popularly known as a diesel generator (DG) for isolated areas
rectifier and a boost converter to reduce the control that only have best solar potential [22]. In this MG, the DG
complexity, sensors requirement and the cost. All these energy has to run every day from evening to morning including the
sources are connected to AC bus using a VSC and a night and during cloudy days. Therefore, small rating BESS is
transformer. The transformer is used for galvanic isolation of required for backup support during change over or during the
DGUs with rest of the system and has delta-star configuration periods when DG is lightly loaded. For small scale MG,
to create four-wire distribution system. All the loads are variable speed DGs shown in Figs. 2 (c)-(e), can be used with
supplied from AC bus known as PCC. a DBR (Diode Bridge Rectifier) and a boost converter to
These MGs are also reinforced with BESS to compensate connect it to DC link, similar to MGs shown in Figs.1 (c)-(e).
the output power fluctuations of DGUs. A dump load is
connected on DC bus and controlled to protect the batteries
from overcharging and to avoid the impact of the dump load
on the power quality of AC bus. DG is attached as an
emergency ES in all these MGs. It operates only if the load
power demand is higher than total generated power of all
DGUs in that MG. It is used to supply load and to charge
BESS, simultaneously. The VSC is controlled for stable AC
bus voltage and constant frequency along with improved
power quality under all load conditions. The boost converter
or VSCs, connected to PV and WT are controlled for MPPT.
A. MG Configurations Based on Two Energy Sources
There are locations, which have two RESs, available
throughout the year. However, some locations have only one
type of RESs available; therefore, such areas require DE
(Diesel Engine) based generator for backup. Such MGs based
on two ESs, consisting of WT, MHP, DG or PV, using various Fig. 2 MG configurations based on PV and DE driven: a) SG, b) SyRG, c)
PMBLDCG, d) SCIG, e)PMSG, and f)DFIG
VSGs, are presented here.
3) Configurations Based on PV and MHP
1) Configurations Based on PV and WT Fig.3 shows MGs based on PV and MHP driven VSG
Fig.1 shows MGs based on PV and WT driven VSG, aimed for isolated areas that have best PV potential and
proposed for isolated areas that have a good wind and solar flowing water all the year [23-24]. In this case, a small hydro
potential [21]. This MG requires proper wind speed to meet reservoir is created so that a MHP generator can run as a
load demand from evening, throughout night and in morning motor in place of a dump load to pump back the water during
till PV power is sufficient enough for loads, otherwise, a large day time when PV power is in excess of load demand and use
BESS is required for this MG to ensure reliable power. For this stored water in evening and night time for MHP generator.
small-scale MG, as shown in Figs.1 (c)-(e), WTs are preferred Otherwise, additional BESS is required as per load demand
to connect at DC bus using a DBR and a boost converter. This and maximum duration for which PV and MHP are not
reduces control complexity and sensor requirements. available. Figs.3 (b)-(c), are advocated for small scale
installation, they are simple and cost effective. Other

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
3

configurations, are preferred for large scale due to the rating of generator is run as a motor in place of a dump load for
electrical machines. pumped storage reservoir during the day time when the WT
power is in excess of load demand for use in the evening and
night time else additional BESS installation is required as per
load demand and duration of non-availability of both RESs.

5) Configurations Based on MHP and DG

The MGs based on variable speed MHPs and DGs, as shown
in Fig.5, are proposed for remote areas having local river
tributaries with high risk of dry season [26]. In Fig.5 (f), DFIG
is proposed as a generator for MHP in order to reduce the
rating of VSC for large scale application. In MGs shown in
Figs.5 (c)-(e), DE based PMBLDCG and PMSG are connected
to DC bus for small scale application through a DBR and a
boost converter in order to reduce the system complexity and
cost. The operation of DG is required only when MHP is not
sufficient enough for load demand. BESS capacity is required
for backup support during change over or to avoid the DG
running on light load for good system efficiency. The MG
shown in Fig.5 (a) is complex and costly; however, it ensures
the power even during faults at VSC level as compared to
Fig. 3 MG configurations based on PV and MHP driven: a) SG, b) SyRG, c) other MG presented in Fig.5. From cost and simplicity point
PMBLDCG, d) SCIG, e) PMSG, and f) DFIG
of view, MGs shown in Figs.5 (b)-(c), are preferable choices.
4) Configurations Based on WT and MHP
Fig.4 shows MGs based on variable speed WT and MHP [25].
These MGs are proposed for isolated areas that have excellent
yearly wind potential and are close to rivers. As presented in
Fig.4, SCIG is used for MHP because it generates fixed
frequency irrespective of loading condition and excitation
capacitance, however, SG, PMSG, SyRG, PMBLDCG, and
DFIG can also be used after evaluation of control
requirements. To avoid a synchronization issue between
DGUs and AC bus, all generators are connected to DC side
except MG shown in Fig.4 (f) for variable speed DFIG where
stator terminals are tied directly to AC side. Amongst the MGs
of Fig.4, a configuration in Fig.4 (f), is preferred for large
scale power generation where MHP and DFIG are used for
cost effective solution. However, for small Fig. 5 MG configurations based on MHP driven SyRG and DE driven :a) SG,
b) SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and in f) DFIG is used for
MHP and DE

6) Configurations Based on WT and DG

The MGs based on variable speed WT and DG are shown in
Fig.6, as proposed for remote areas that possess excellent
resource of wind [10, 27]. In this MG, DG is operated only
when WT power is not sufficient enough for load demand.
During change over from RES to DG, BESS is required as
backup support. This also avoids DG running at light load to
have good efficiency. As shown in Fig.6, all WTs are having
variable speed SCIGs, which can be replaced by SyRG, SG,
PMSG, PMBLDCG, and DFIG along with required control.
For cost effectiveness and simplicity, MGs shown in Figs.6
(b)-(e), are the best options. For large rating applications, MG
shown in Fig.6 (f), is preferred.
Fig. 4 MG configurations based on MHP driven SCIG and WT and driven: a) B. MG Configurations Based on Three Energy Sources
SG, b) SyRG, c) PMBLDCG, d) SCIG, e)PMSG, and f)DFIG
To take care an uncertain availability of RESs, the energy
scale installation MHP driven fixed speed generator is simple sources can be increased. All such MGs having any three
and cost effective solution. In this MG also, the MHP

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
4

DGUs from WT, MHP, DG and PV, are proposed. These MGs In these MGs, DG has to be used for critical loads only to
are used to ensure stable and interrupted power supply to the avoid high installation cost of MG. Running of MHP as a
consumers under severe weather conditions. motor for pumped storage is also an option if possible. Small
rating BESS is required to support critical loads during change
over between ESs. MGs shown in Figs.8 (b)-(c), are proposed
for small power applications whereas, for medium and large
rating applications, any MG shown in Fig.8, is preferred.

Fig. 6 MG configurations based on WT driven SCIG and DE driven :a) SG, b)

SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and in f) DFIG is used for MHP
and DE

1) Configurations Based on PV, WT and MHP

In Fig.7, MGs are presented for isolated areas that have solar
and wind potential near a canal or run-of-river [28]. In these
MGs, a pumped storage plant or a large BESS is essentially
required to handle eventualities due to non-availability of
RESs. For small and medium scale installations, MGs shown
in Figs.7 (b)-(e) are suggested and for large scale installation,
Fig. 8 MG configurations based on DE driven SG, MHP driven SyRG and
MGs shown in Figs.7 (a)-(e) are preferred. For small power WT driven: a) SG, b) SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and f) DFIG
applications, it is preferable to use a DBR and a boost
converter to tie WTs with MGs shown in Figs.7(c) and (e). 3) Configurations Based on PV and MHP and DG
The MGs based on PV, DG and MHP are shown in Fig.9
[30]. These MGs are proposed for remote areas that possess
excellent resource of PV and flowing water. The DG capacity
is selected only for critical loads in these MGs to avoid high
total installation cost. The generator used for MHP is run as a
motor as a pumped storage, if possible, to have a buffer RES.
Small rating BESS is sufficient to support critical loads during
change over between ESs. For large power applications, the
MGs shown in Figs.9 (f), is preferred and for small scale, as
well as medium ratings, MG shown in Figs.9 (a)-(e), are
recommended.
4) Configurations Based on PV and WT and DG
In Fig.10, the MGs are presented for isolated areas that have
variable solar and wind potential [31]. In these MGs, the DG
is used as back up ES only to supply the critical load, else the
installation cost of this configuration is very high. Small rating
BESS is sufficient for this configuration to support during
change over. For large rating applications, MG shown in
Figs.10 (a) and (f), are preferable. Other configurations are
Fig. 7 MG configurations based on PV, and MHP driven SCIG, and WT best choice for small and medium power applications.
driven: a) SG, b) SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and f) DFIG
C. MG Configurations Based on Four Energy Sources
2) Configurations Based on MHP, DG and WT In Fig.11, MGs based on PV and variable speed DG, MHP
MGs based on MHP, DG and WT shown in Fig. 8, are and WT are presented [32]. These MGs are proposed for
aimed to isolated areas that have best wind potential and remote areas that possess good solar, wind potential and
situated near rivers where a risk of dry season is possible [29]. located closer to rivers. These MGs are designed in such a

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
5

way that the use of BESS and DG is only for critical loads or
during change over between various ESs. For large power
applications, the MGs shown in Figs.11 (a)-(f) can be used,
and for small and medium rating applications, MGs shown in
Figs.11 (b)-(e) are suggested as good alternative.

Fig. 11 MG configurations based on PV, DE driven SG, MHP and WT driven

a) SG a SCIG, b) SyRGs, c) PMBLDCGs, d) SCIGs, e) PMSGs, and f)
DFIGs

III. CONTROL APPROACHES FOR MG

Various control approaches have been reported in the
literature for MPPT from PV, WT, MHP and DG [33-38], to
Fig. 9 MG configurations based on PV, DG driven SG, and MHP driven a)
SG, b) SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and f) DFIG
control BESS charging, DC dump load and power
management in MG [39-40], to control frequency and voltage,
synchronize ESs, and to improve power quality at PCC [41].
For MGs presented here, these controllers are summarized in
four broad categories as controllers for boost converters of
PV, for VSC of variable speed WT, DG and MHP, for DC
dump load control and BESS protection, and for voltage,
frequency regulation and power quality improvement at PCC.
The details of these control approaches are presented here.
A. Control Approach for Boost Converter of PV Array
As shown in all these MGs, the PV array is attached on DC
side through a boost converter, which steps up the output
voltage of PV while achieving MPPT. Among the methods
proposed in the literature, perturbation and observation (P&O)
method is frequently used in practice due to its simplicity.
However, this method suffers under rapid irradiation changes.
This issue is solved using adaptive P&O, combined global
search algorithm with P&O and sliding mode approach with
boundary layer [42-43].
For this control strategy shown in Fig.12, the boost converter
output voltage (vout) and input voltage (vpv), and the inductor
current (iL), which is equal to the output PV current (ipv), are
Fig. 10 MG configurations based on PV, DE driven SG, and WT driven a) required to define the control law (d) obtained as sum of
SG, b) SyRG, c) PMBLDCG, d) SCIG, e) PMSG, and f) DFIG equivalent control (deq) and switching control (ds). The
Further MGs having more than four energy sources may maximum PV current (impv), which represents the reference
also be considered based on the available RESs in a particular PV current is obtained using P&O technique.
region such as, GS, TG, TWG, or BEG. The choice of B. Control Approach for DC Dump Load and BESS
generator and control is based on amount of power available,
As shown in Figs. 1-11, proposed MGs are reinforced by
cost and complexity constraints and discretion of the user.
DC dump load to protect the BESS from overcharging. This

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
6

dump load is tied to the DC side instead of AC side in all MGs r

r Actuator Engine
for economy and prevention of any power quality issues on - -sτ
AC side. It is used as smart load [44], pumping load [45] or + PI k1 / (1+ sτ1) k 2e 2 PMSG
removed completely from these MGs for some applications  PL  PRES PL
[46] as and when desired by the user. rref  PRES
SOC%  50% SOC%
However, to protect BESS from overcharging, the rref PDG
r  f  PDG 
developed control approaches for MPPTs from RESs are ref
designed to work according to load power demand when Teref i q_ref
PI 4Teref (3Pλm )
BESSs are fully charged. For a dump load, reference battery
PWM PI
voltage, which represents maximum battery voltage, is i DG
compared with sensed battery voltage, and the error is fed to + L DG
v +
battery voltage regulator, and its output is compared with dc Cdc SDG v DG
- -
triangular carrier to get duty cycle for switching a dump load.
Fig. 13 Model of DE and the control algorithm for DC-DC boost converter for
DE driven variable speed PMSG

D. Control Strategy for DC-DC Boost Converter of WT Side

WT driven variable speed PMBLDCG supported by BESS
behaves similar to PV array. To achieve MPPT from WT, only
Fig. 12 Improved MPPT based P&O method with sliding mode control output DC current and voltage are required, using power ratio
variable step based P&O method [48] through a boost
C. Control Approach for VSC of Variable Speed DGUs
converter, as shown in Fig.14.
The DEs driven VSGs in proposed MGs are tied to the AC
The output WT power at nth and (n-1)th instant are given as,
and DC both sides of system through VSCs and through DBR-
boost converter. This configuration is used for small scale pWT ( n )  vWT ( n ) iWT ( n ) (4)
applications when either PMBLDCG or PMSG is used as pWT ( n 1)  vWT ( n 1) iWT ( n 1) (5)
generator with DE. Details about the mathematical model of where vWT, and iWT denote the instantaneous values of sensed
DE and its control algorithm are given in [47]. Fig.13 presents
DC output voltage and current of the boost converter for WT.
a model of DE and a control strategy for a boost converter.
This model of DG has two parts namely mechanical parts The differential DC voltage ΔvWT and WT power ΔpWT, is as,
consisting of speed controller, actuator and DE; and a  vWT ( n )  vWT ( n )  vWT ( n 1) (6)
generator as electrical part, which may be either PMSG (as  PWT ( n )  PWT ( n )  PWT ( n 1) (7)
shown in Fig.13) or any other electrical machines as shown in
The sign of factor k, is calculated as,
MGs in Figs.1-11. As already indicated, DG is used as an
emergency ES in all these MGs and operates only if a state of  k   1 if  PWT ( n )  0,
 k  1 if (8)
charge (SOC) of BESS is less than 50% and the generated   PWT ( n )  0
power from RESs (PRESs) is less than load demand (PL) as,
PL  PRESs and SOC %  50% (1) The duty cycle for the next perturbation is calculated as,
The output DG power (PDG) is used to estimate reference d n  d ( n 1)  k  d (9)
rotor speed ωrref using a real power-speed curve given in [32]. where Δd denotes the duty cycle perturbation step for next
The developed algorithm to control a boost converter for DG
perturbation and is calculated as,
is shown in Fig.13. The PDG is calculated as,
d  S * T (10)
PDG  v DG iDG (2)
where ΔT is a fixed step size and S is variable power ratio
where PDG, vDG, iDG denote pervious output DG instantaneous
calculated as,
power, sensed DC output voltage, and current for DG side.
The estimated rotor speed ωrref is compared with sensed S   PWT max  PWT ( n )  / PWT ( n ) (11)
rotor speed ωr, and an error is processed through a PI where PWTmax denotes the maximum WT output power.
controller to get torque reference (Teref). From this, q-axis E. Control Strategy for AC-DC Converter for MHP
reference current (iqref), is calculated as,
Most of the MHP are run of river type, which possess low
iqref  (4 / 3)(Teref / P  m ) (3) heads with high water flow rates. Kaplan and semi-Kaplan
turbines are good choice as prime movers for this application
where, P and λm represent number of pole and flux,.
[30]. Regarding the electrical machine, classical as well as
The error between iqref and iDG is fed to PI current controller. special machines can be used as a generator. For MPPT from
The PI controller output generates switching signal (SDG) for MHP, the available techniques of PV and WT can be applied
the boost converter using a PWM controller. with some modifications, if required [18].

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
7

i WT Fig.16 shows the detailed scheme of the selected MG

+ SWT L WT
+ configuration based on WT driven PMBLDCG with PV
v dc Cdc v WT a
- -
b c
ωr
(Fig.16a) and with DE driven PMSG (Fig.16b).
vWT(n)
i WT(n)
PM
BLDCG
P
WTmax -PWT(n) 
PWT(n)
Z-1 Z-1
PWT(n) PWT(n-1)
-
PWT(n)
WT
S d +

T
If PWT(n)  0
k = -1
d(n) kd k
SWT +
+
k =1
Z-1 d (n 1)
If p WT(n)  0

Fig. 14 Control algorithm of boost converter for variable speed WT driven

BLDCG
The available hydraulic power (Ph) is function of the water
head (H), and the water flow (Qω) given as, Fig. 16a Detailed scheme of the MG configuration based on WT and PV
P   gHQ (12)
h 
where ρ and g denote the volume density of water (kg/m3) and
the acceleration of water (m/s2), respectively.
The extractible mechanical power (PT) is defined as,
P  P (13)
T h
where η represents the hydraulic turbine efficiency.
The hydraulic turbine efficiency [22] is expressed as,
   ,Q   0.5 90 i Q 0.78 e 50 i    3.33Q  (14)
 

 
Fig. 16b Detailed scheme of the MG configuration based on DG and WT
 i  1 /  1   0.089  0.0035 and   RA Q (15)
where R, A, Ω denote the radius of the turbine blades, area Applying Kirchhoff’s voltage and current laws to the VSC and
swept, and the rotation-speed of the turbine, respectively. the load terminal in Fig.16, resultant equations are given as;
F. Control Approach for Voltage and Frequency Regulation diinvabc 1
and Power Quality Improvement    v Labc  d abc v dc  (16)
dt Lf
In all MGs shown in Figs. 1-11, a VSC is operated to
dv Labc 1
control the voltage and frequency at PCC with balancing of   iinva bc  iLabc  (17)
load and harmonics mitigation. Many techniques reported in dt Cf
the literature to realize these tasks, include synchronous where, dabc, vLabc, iinvabc, Cf, Lf represent the controls laws, AC
reference frame technique, nonlinear control technique, load voltages, output inverter currents, capacitance of the
instantaneous power theory, instantaneous symmetrical output filter and output inverter inductance, respectively.
component theory and modified Fortescue's theory [49]. Some Replacing (16) in the derivative of (17), results as,
MGs do not require synchronization between DGUs and PCC d 2 v Labc 1  1 di 
as all DGUs are tied to the DC side, thereby avoiding the 2
    v Labc  d abc vdc   Labc  (18)
d t C f  Lf dt 
phase locked loop (PLL) or measurement of system frequency
at PCC. To maintain constant system frequency, it (fs) is fixed Applying Park’s transformation to (18), gives equations as,
at 60Hz. Fig.15 shows a scheme for AC voltage regulation  d 2v  1  v 1 di Ld
 2 Ld      2  v Ld  dc d d 

based on sliding mode control.  d t C L C L C dt
 f f  f f f
 dv Lq
  2 dt (19)
 2
 d v Lq  1 
2  v 1 di Lq
 d t 2
 
 C L
  v  dc d 
 Lq C L q C dt
  f f  f f f
dv
  2 Ld
 dt
The sliding mode control for AC load voltage regulation is
obtained using the following steps.
1) Choice of sliding surface
The sliding surface in d-q axis σd and σq, is defined as [43],

Fig. 15 Control scheme for AC voltage and frequency regulation.

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
8

  d    1  1 diLd dv Lq 
 d   dt  K 1   v Ld (20)      2  v Ld   2 
  d  f f
C L
 C f L f  C f dt dt 
 
 d
 q    K 1   v Lq  eqd
v dc  d  v Ld  
  dt     K1  (28)
  dt 
where K1 is positive gain and ΔvLd, ΔvLd are the load voltage 
  1  1 diLq dv 
errors which are given as,     2  v Lq   2  Ld 
 C f Lf C L  C dt dt
  v Ld  v Ld  v Ld ref (21)  d eqq   f f  f 
 v  v  v  v dc  d  v Lq  
 Lq Lq Lq ref
   K1 
where vLdref is desired direct AC load voltage. The voltage in   dt 
quadrature vLqref is kept equal to zero.
2) Determination of the equivalent control IV. SIMULATION AND EXPERIMENTAL RESULTS
To regulate constant AC load voltage, following conditions To demonstrate operation of proposed MGs in terms of
are imposed: desired DGU characteristics such as MPPT of PV array,
 v Ldref  v LL frequency and voltage control at PCC, ensuring continuous
(22)
 v Lqref  0 power to loads, power quality improvement at PCC and
    t  2.. f s .t overcharging protection of BESS, two MGs based on two
where vLL, fs denote the line-line AC load voltage and the energy sources namely PV-WT and DG-WT, are modeled and
system frequency, respectively. simulated in Matlab/Simulink. The details of simulation
The control input is defined as, parameters are summarized in Tables I and II.
TABLE I
 d d  deq d  k 2 sgn(  d ) (23) PARAMETERS OF MG CONFIGURATION BASED ON PV-WT

 d q  deq q  k 2 sgn(  q ) Element Parameters
where k2 is a positive gain. PV array 8
irr=5.981. 10 A, iscr=6 A, ki=0.0024, Tr=298 K
The equivalent control is obtained from the invariance And DC-DC q=1.6. 1019 C, Kb=1.38. 1023 J/K, Eg=1.12 V, A=1.2
conditions given as, Boost1 Cout1=1000µF, C1=1000 µF, L1=1.5 mH, k1=50
 d d WT Cpmax=0.48, λopt=8.1, C1=0.5176, C2=116, C3=0.4, C4=5,
 d  0  dt  0  d d  deq d (24)
C5=21, C6=0.0068
 PMBLDCG Rs=0.808 Ω, Ls=5.44 mH, Vs=208 V, ωr= 1800 RPM,
d q And DC-DC J=0.01859 kg.cm2, Km=80 V/tr/min, Cout2=1000µF, C2=1000
 q  0   0  d q  deq q
 dt Boost2 µF, LWT=1.5 mH, 2P=4, k2=50
Bus DC Cdc=2500 µF, Vdc=105 V, lead acid batteries 9*(12V/12) Ah.
And the derivative of (20) is given as,
AC local fs=60 Hz, VLL=50V, Lf=5mH, Cf= 40 µF linear load (RL=8
 d d d 2  v Ld d  v Ld grid Ω), nonlinear load ( RL=8 Ω, LL=20 mH ), k1=k2=0.01
   K1
dt 2
d t dt (25)
 d d 2
 v d  v Lq TABLE II
 q

Lq
 K1 PARAMETERS OF MG CONFIGURATION BASED ON DG-WT
 dt 2
d t dt
Element Parameters
Replacing (21) in (25) gives;
 
d 2 v Ld  v Ld ref 
d v Ld   v Ld ref  DE model &
PMSG
Ki=120, kp=10, k1=1, k2=1, τ1=0.4, τ2=0.011, Rs=0.005 Ω,
L=0.000835H, flux linkage=0.5V.s, J=0.25kg.m2,
0   K1
 2
dt (26)
   
d t F=0.005N.m.s
 d 2
v  v Lq ref d v Lq  v Lq ref WT Cpmax=0.48, λopt=8.1, C1=0.5176, C2=116, C3=0.4, C4=5,
0  Lq
 K &PMBLDCG C5=21, C6=0.0068, Rs=1.085 Ω, L=3.305mH, flux
 d 2t
1
dt linkage=1.1233V.s, J=0.0068kg.m2, F=0.0001021N.m.s
BESS and Vbat=1000V, Vocmin=980V, Vocmax=1088V, Cb=4560F ,
substituting (26) in (19), results as; dump load Rs=0.01 Ω, Rb=10k Ω, Rd=97.08 Ω
  1  v 1 di Ld AC local VLL=460V, fs=60Hz, R=10 Ω, nonlinear load (Diode
 0      2  v Ld  dc d eqd 
 grid bridge), RL=50 Ω, LL=60e-3mH
  C f Lf  C f Lf C f dt
 dv Lq d v Ld 
  2 dt  K 1 dt (27) Fig.17 shows a hardware used to validate selected
 configurations. It consists of: 1) four-quadrant dynamometer/
 0    1   2  v  v dc d  1 di Lq power supply, 2) PMBLDC generator, 3) a drive, with 4)
  C L  Lq C L eqq C dt
 
d v Lq 
 f f f f f squirrel cage induction machine and 5) synchronous machine,
dv 6) lead acid battery pack, 7) power converters, 8) voltage and
  2 Ld  K 1
 dt dt currents sensors, 9) transformer, 10) DSP controller and
Arranging (27), gives the control laws in d-q axis as, protection cards, 11) PV emulator, and 12) loads.

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
9

(a) (b)

Fig. 17 Hardware setup for MG realization

Fig.18 (a) shows waveforms of the output PV voltage (Vpv),
and current (iPV), the inductor current (iL1) and the reference
PV current (imPV) under solar irradiance change. It is observed
that iPV varies with variation of the solar irradiation. One can
see clearly that the improved P&O based on sliding mode
control perform well during sudden increasing and deceasing
of solar irradiance where iPV follows its reference imPV. (c) (d)
Fig.19 Simulated performance: a) steady state, b) load change of phase (a), c)
In Fig.18 (b), waveforms under wind speed change of the sudden increased of load, and d) balanced and unbalanced nonlinear load.
mechanical torque (Tm), rotor speed of the PMBLDCG (ωr),
terminal stator voltage (Vsa) and current (isa) of the phase (a),
the DC voltage (VWT) and current (iWT), are presented. It is
observed that improved P&O perform without rotor speed
sensing during wind speed change. The iWT follows variation
of (isa), it increases and decreases with variation of Tm but
the VWT varies slightly with this variation, which help to
reduce the complexity and the cost of the system by using
only the sensed VWT and iWT to achieve MPP from WT.
0
[N.m]
Insolation

1000
[W/m²]

-20
Tm

-40
0 0.4 0.9 1.4 1.7
60
500
0 0.3 0.8 1.3 1.7
[rad/s]

50
wr

60
40
55 0 0.4 0.9 1.4 1.7
50
[V]
vpv

[V]

0
vsa

50
-50
45
0 0.3 0.8 1.3 1.7 0 0.4 0.9 1.4 1.7
5
5
[A]

0
isa
[A]

-5
ipv

0 0.4 0.9 1.4 1.7

70
0 60
[V]
vWT

0 0.3 0.8 1.3 1.7

5 50
0 0.4 0.9 1.4 1.7
5
iL1&impv
[A]

[A]
iWT

0 0
0 0.3 0.8 1.3 1.7 0 0.4 0.9 1.4 1.7
(a) Time [s] (b) Fig.20 Experimental results of nonlinear control based on sliding mode
Fig.18 Performance of WT and PV at change of: a) solar irradiance and b) control under: a) steadystate, b) sudden increasing in load, c) sudden
wind speed. switching off phase a, d) no load, and e)balanced nonlinear load

In Fig.19 (a-d), waveforms of DC link voltage (vdc), which Experimental performance of this MG is shown in Figs.20-22.
actually represents a battery voltage (vbat), load voltage (vL) Figs.20 (a)-(b) demonstrate that AC voltage is regulated
and current (iL), and the frequency at the PCC. It is observed constant and sinusoidal during sudden load variation and when
that proposed control strategy based on sliding mode control nonlinear load is connected to PCC. In Fig.21, performance of
shown in Fig.15 performs well during presence of severe improved P&O MPPT is demonstrated under wind speed
conditions such as, load variation when one phase is switched change, as the output WT current variation. In Fig.22 (a)-(d),
off (Fig.19b), sudden increasing in load (Fig.19c) and during performance of proposed MG at varying solar irradiance as
presence of balanced and unbalanced nonlinear load (Fig.19d). well as wind speed change are demonstrated. It is observed
It can be seen that vL is regulated constant and sinusoidal. that AC voltage is regulated constant and sinusoidal during

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
10

solar irradiation and wind speed. The battery current is 375

[V]
L
0
observed with variation of wind speed and solar irradiation in

V
-375
0 0.3 0.5 1.1 1.7
order to balance power in the system. 20

[A]
0

L
i
-20
0 0.3 0.5 1.1 1.7
60.5

[Hz]
60

s
f
59.5
0 0.3 0.5 1.1 1.7

WT
[m/s]
10

V
5
0 0.3 0.5 1.1 1.7
0

[N.m]
m
-100

T
-200
0 0.3 0.5 1.1 1.7
20

WT
[A]
0

i
-20
0 0.3 0.5 1.1 1.7
15

[A]
dcWT
10
5
0

i
0 0.3 0.5 1.1 1.7
50

[rad/s]
DG
w&
0

r
w
-50
0 0.3 0.5 1.1 1.7
10

dcDG
[A]
0
-10

i
0 0.3 0.5 1.1 1.7
1085
Fig. 21 a) Experimental results of the improved P&O for MPPT from WT

dc
[V]
V
driven variable speed PMBLDC generator, and b) zoom of (a) 1080
0 0.3 0.5 1.1 1.7

SOC%
69.999
69.9985
69.998
0 0.3 0.5 1.1 1.7
Time [s]

Fig.23 Simulated performance of WT-DG MG under load and wind speed

change when state of charge of battery is greater than 50%
In Fig.24, the simulated dynamic performance of MG based
on DE driven variable speed PMSG and WT driven variable
speed PMBLDCG are presented when SOC is less than 50%.
375
[V]
L

0
V

-375
0 0.3 0.5 1 1.5
20
[A]

0
L
i

-20
0 0.3 0.5 1 1.5
[Hz]

60
s
f

59.9
00 0.3 0.5 1 1.5
[N.m]
m
T

-25
0 0.3 0.5 1 1.5
2
WT
[A]

0
-2
i

0 0.3 0.5 1 1.5

5
dcWT
[A]

0
i

0 0.3 0.5 1 1.5

[rad/s]

150
DG
w&

Fig.22 Experimental results of improved P&O based on sliding mode control 140
w
r

130
with boundry layer for WT and PV sides at a) sudden increasing wind speed, 0 0.3 0.5 1 1.5
b)sudden decreasing of wind speed, c) sudden decreasing of solar irradiance,
dcDG

10
[A]

and d) ) sudden decreasing of solar insolation with fixed speed wind speed. 5
i

0 0.3 0.5 1 1.5

1080
1078
dc
[V]

The dynamic performance of variable speed DE driven

V

1076
0 0.3 0.5 1 1.5
PMBLDCG and WT driven variable speed PMSG, are 49.504
SOC%

49.502
demonstrated in Figs 23-24 at varying load, wind speed and 49.5
0 0.3 0.5 1 1.5
state of charge of battery. Time [s]

Fig.23 shows AC voltages (vL), load currents (iL), system Fig. 24 Simulated performance of WT-DG MG under load variation when
state of charge of battery is less than 50%
frequency (fs), wind speed (VWT), torque of WT (Tm), stator
currents of PMBLDCG (iWT), output current of boost It is observed that load current is greater than WT current
converter WT side (idcWT), rotor speed of PMSG and rotor and SOC% is less than 50%, therefore DG is operating.
speed of DE (wr & wDG), output current of boost converter DG Further, the rotor speed of PMSG follows its reference, which
side, DC link voltage, and state of charge of battery (SOC%). represents DE speed and the battery is being charged while
These results are shown at varying load and wind speed with supplying the load. The presented simulated performances and
SOC% greater than 50%. It is observed that DG is not in test results have validated proposed control concepts for
operation whereas WT and BESS are supplying the load. The presented MG configurations while ensuring continuous
stator currents iWT, and idcWT vary with varying wind speed power to various loads in isolated and remote areas.
shown as increment at t=0.3s and at 1.1s. The AC voltage and
frequency remain constant during load and wind speed V. APPLICATIONS
changes, which confirms robustness of control. In remote and isolated areas or hilly regions, the power
generating systems using any of these micro-grids are required

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
11

for applications like home lighting, appliances, street lighting, [13] S. M. Malik, X. Ai, Y. Sun, C. Zhengqi, and Z. Shupeng, “Voltage and
frequency control strategies of hybrid AC/DC microgrid: a review,” IET
community water supply pumps, water treatment plants for
Gen. Trans. & Distr., vol.11, no.2, pp.303-313, 2017.
drinking water supply, waste water treatment and disposal [14] K. Kant, C. Jain, and B. Singh, “A Hybrid Diesel-Wind-PV based
system, pastoral stations or cattle farming (i.e. pig, sheep or Energy Generation System with Brushless Generators,” IEEE Trans.
goat farming), chicken farming, dairies, tourist facilities such Industrial Informatics, vol. 13, no. 4, pp. 1714-1722, 2017.
[15] N.Mendis, K.M.Muttaqi, S.Perera and M.N.Uddin, “Remote Area Power
as cottage and resorts, aerial ropeway transport system,
Supply System: An Integrated Control Approach Based on Active Power
restaurant cum recreation centers, mine sites and military Balance,” IEEE Ind. Appl. Mag., vol.21, no.2, pp.63-76, Apr.2015.
bases. Other applications include data centers, telecom towers [16] M. Rezkallah, S. Sharma, A. Chandra, and B. Singh, “Hybrid standalone
and exchanges, petroleum refilling stations, hospitals and power generation system using hydro-PV-battery for residential green
buildings,” in Proc.IECON 2015, pp. 3708-3713.
nursing homes, day care centers, old age homes, EV battery
[17] K. M. Reddy and B. Singh, “Dual Mode Multi-functional Small Hydro
charging centers on highways and rural irrigation system. and SPV Generation Based Reconfigurable System under Non-Ideal
Grid Conditions," IEEE Trans. Smart Grid, Early Access, 2017.
VI. CONCLUSIONS [18] L. Belhadji, S. Bacha, I. Munteanu, A. Rumeau, and D. Roye, “Adaptive
MPPT Applied to Variable-Speed Microhydropower Plant,” IEEE
A comprehensive study of various configurations of micro- Trans. Energy Con., vol. 28, no. 1, pp. 34-43, 2013.
grid systems has been carried out. These MGs are classified [19] J. W. Simpson-Porco, Q. Shafiee, F. Dörfler, J. C. Vasquez, J. M.
Guerrero, and F. Bullo, “Secondary Frequency and Voltage Control of
and presented on the basis of number of energy sources along
Islanded Microgrids via Distributed Averaging,” IEEE Trans. Industrial
with their control, selection criteria, performance simulation Elect., vol. 62, no. 11, pp. 7025-7038, Nov. 2015.
and experimental validation. The presented simulated [20] W. Kohn, Z. B. Zabinsky, and A. Nerode, “A Micro-Grid Distributed
performance and its experimental validation on prototypes, Intelligent Control and Management System,” IEEE Trans. Smart Grid,
verify the applications of MG configurations at desired vol. 6, no. 6, pp. 2964-2974, Nov. 2015.
[21] M.B. Shadmand and R.S. Balog, "Multi-Objective Optimization and
conditions. These configurations, control algorithms and Design of Photovoltaic-Wind Hybrid System for Community Smart DC
applications are expected to augment new ideas for extraction Microgrid," IEEE Trans. Smart Grid, vol.5, no.5, pp.2635-2643, Sep.
of various renewable energy sources and their applications for 2014.
sustainable development in rural and isolated regions [22] M. Datta, T. Senjyu, A. Yona, T. Funabashi, and C.H. Kim, "A
Frequency-Control Approach by Photovoltaic Generator in a PV-Diesel
REFERENCES Hybrid Power System," IEEE Trans. Energy Con., vol. 26, no2, pp. 559-
571, June 2011.
[1] V. Yaramasu, B. Wu, P. C. Sen, S. Kouro, and M. Narimani, “High- [23] R. Joseph and L. Umanand, "A Brushless Wound Rotor Induction
power wind energy conversion systems: State-of-the-art and emerging Generator for Variable Speed Microhydel Plants without Ballast Load,”
technologies,” IEEE Proc., vol.103, no.5, pp.740-788, May 2015. IEEE Trans. Sust. Energy, vol. 6, no1, pp. 20-27, Jan 2015.
[2] M. F. M. Arani and Y. A. R. I. Mohamed, “Dynamic Droop Control for [24] I. Tamrakar, L. B. Shilpakar, B. G. Fernandes, and R. Nilsen, “Voltage
Wind Turbines Participating in Primary Frequency Regulation in and frequency control of parallel operated synchronous generator and
Microgrids,” IEEE Trans. Smart Grid, Early Access, 2017. induction generator with STATCOM in micro hydro scheme, ” IET
[3] O. D. Mipoung, L. A. C. Lopes, and P. Pillay, “Potential of Type-1 GTD, vol. 1, no5, pp. 743-750, Sept. 2007.
Wind Turbines for Assisting With Frequency Support in Storage-Less [25] A. T. Thankappan, S. P. Simon, P. S. R. Nayak, K. Sundareswaran, and
Diesel Hybrid Mini-Grids,” IEEE Trans. Ind. Electron., vol. 61, no. 5, N. P. Padhy, “Pico-hydel hybrid power generation system with an open
pp. 2297-2306, May 2014. well energy storage, ” IET GTD, vol. 11, no3, pp. 740-749, Feb.2017.
[4] B. Singh, R. Niwas, and S. K. Dube, “Load Leveling and Voltage [26] M. F. Alkababjie and W. H. Hamdon, “Feasibility and environmental
Control of Permanent Magnet Synchronous Generator-Based DG Set for effects study of adding micro hydro power plant, converter and batteries
Standalone Supply System,” IEEE Trans. Ind. Infor., vol. 10, no.4, pp. to diesel generators using in electrification a remote Iraqi village,” in
2034-2043, Nov. 2014. Proc. FNCES 2012, pp. 1-6.
[5] B. Singh and S. Sharma, “Stand-Alone Single-Phase Power Generation [27] R. Sebastián, “Battery energy storage for increasing stability and
Employing a Three-Phase Isolated Asynchronous Generator,” IEEE reliability of an isolated Wind Diesel power system,” IET Ren. Power
Trans. Industry Applns, vol. 48, no. 6, pp. 2414-2423, Dec. 2012. Generation, vol. 11, no2, pp. 296-303, Feb. 2017.
[6] B.Singh and V.Rajagopal, “Neural-Network-Based Integrated Electronic [28] A. K. Bansal, R. Kumar, and R. A. Gupta, “Economic Analysis and
Load Controller for Isolated Asynchronous Generators in Small Hydro Power Management of a Small Autonomous Hybrid Power System
Generation,” IEEE Trans. Ind. Elect., vol.58, no.9, pp.4264-4274, (SAHPS) Using Biogeography Based Optimization (BBO) Algorithm, ”
Sep.2011. IEEE Trans. Smart Grid, vol. 4, no1, pp. 638-648, March 2013.
[7] M. A. Tankari, M. B. Camara, B. Dakyo, and G. Lefebvre, “Use of [29] R. C. Bansal, T. S. Bhatti, and D. P. Kothari, “Automatic reactive power
Ultracapacitors and Batteries for Efficient Energy Management in Wind- control of wind-diesel-micro-hydro autonomous hybrid power systems
Diesel Hybrid System,” IEEE Trans. Sustainable Energy, vol. 4, no. 2, using ANN tuned static VAr compensator,” in Large Engg. Systems
pp. 414-424, April 2013. Conf. on Power Engg., 2003, pp. 182-188.
[8] P. Sharma and T. S. Bhatti, “Performance Investigation of Isolated [30] K. Kusakana, J. L. Munda, and A. A. Jimoh, “Feasibility study of a
Wind-Diesel Hybrid Power Systems With WECS Having PMIG,” IEEE hybrid PV-micro hydro system for rural electrification,” in Proc.
Trans. Ind. Elect., vol. 60, no. 4, pp. 1630-1637, April 2013. AFRICON, 2009, pp. 1-5.
[9] G. Pathak, B. Singh, and B. K. Panigrahi, “Back-Propagation [31] A. Mohanty, S. Patra, and P. K. Ray, “Robust fuzzy-sliding mode based
Algorithm-Based Controller for Autonomous Wind-DG Microgrid,” UPFC controller for transient stability analysis in autonomous wind-
IEEE Trans. Industry Appl., vol. 52, no. 4, pp. 4408-4415, Dec. 2016. diesel-PV hybrid system,” IET GTD, vol.10, no.5, pp. 1248-1257, Jul.
[10] R. PeÑa, R. CÁrdenas, J. Proboste, J. Clare, and G. Asher, “Wind- 2016.
Diesel Generation Using Doubly Fed Induction Machines,” IEEE Trans. [32] M. M. A. Rahman, A.T.A. Awami, and A.H.M.A. Rahim, “Hydro-PV-
Energy Con., vol. 23, no. 1, pp. 202-214, March 2008. wind-battery-diesel based stand-alone hybrid power system,” in Proc.
[11] T. Dragičević, X. Lu, J. C. Vasquez, and J. M. Guerrero, “DC Int. Conf. Elect. Engg. Inf. & Com. Tech., 2014, pp. 1-6.
Microgrid-Part II: A Review of Power Architectures, Applications, and [33] B. Singh, M. Kandpal, and I. Hussain, “Control of Grid Tied Smart PV-
Standardization Issues,” IEEE Trans. Power Elect., vol. 31, no. 5, pp. DSTATCOM System using an Adaptive Technique,” IEEE Trans.
3528-3549, May 2016. Smart Grid, Early Access, 2016.
[12] J. M. Guerrero, P. C. Loh, T. L. Lee, and M. Chandorkar, “Advanced [34] S. Adhikari and F. Li, “Coordinated V-f and P-Q Control of Solar
Control Architectures for Intelligent Microgrids-Part II: Power Quality, Photovoltaic Generators with MPPT and Battery Storage in Microgrids,”
Energy Storage, and AC/DC Microgrids,” IEEE Trans. Industrial Elect., IEEE Trans. Smart Grid, vol. 5, no3, pp. 1270-1281, May 2014.
vol. 60, no. 4, pp. 1263-1270, April 2013.

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TSG.2017.2762349, IEEE
Transactions on Smart Grid
12

[35] Z. Cui, L. Song, and S. Li, “Maximum Power Point Tracking Strategy
for a New Wind Power System and Its Design Details,” IEEE Trans
Energy Conversion, vol.32, no.3, pp. 1063-1071, 2017.
[36] C. Wei, Z. Zhang, W. Qiao, and L. Qu, “Reinforcement-Learning-Based
Intelligent Maximum Power Point Tracking Control for Wind Energy
Conversion Systems,”IEEE Trans. Ind. Electron., vol. 62, no.10, pp.
6360-6370, Oct.2015.
[37] L. Belhadji, S. Bacha, and D. Roye, “Modeling and control of variable-
speed micro-hydropower plant based on Axial-flow turbine and
permanent magnet synchronous generator (MHPP-PMSG),” in Proc
IECON 2011, pp. 896-901.
[38] J. Fraile-Ardanuy, J.R. Wilhelmi, J.J. Fraile-Mora, and J.I. Perez,
“Variable-speed hydro generation: operational aspects and control,”
IEEE Trans. Energy Con., vol. 21, no2, pp. 569-574, June 2006.
[39] A. Nisar and M.S. Thomas, “Comprehensive Control for Microgrid
Autonomous Operation with Demand Response,” IEEE Trans Smart
Grid, vol.8, no.5, pp. 2081-2089, 2017.
[40] T. Morstyn, B. Hredzak, and V.G. Agelidis, “Control Strategies for
Microgrids with Distributed Energy Storage Systems: An Overview,”
IEEE Trans. Smart Grid, vol. Early access, 2016.
[41] Jianguo Zhou, Sunghyok Kim, Huaguang Zhang, Qiuye Sun and Renke
Han, “Consensus-based Distributed Control for Accurate Reactive,
Harmonic and Imbalance Power Sharing in Microgrids,” IEEE Trans.
Smart Grid, Early Access, 2017.
[42] C. Manickam, G. R. Raman, G. P. Raman, S. I. Ganesan, and C.
Nagamani, “A Hybrid Algorithm for Tracking of GMPP Based on P&O
and PSO With Reduced Power Oscillation in String Inverters, ” IEEE
Trans. Ind. Electron., vol. 63, no.10, pp. 6097-6106, Sept. 2016.
[43] M. Rezkallah, A. Hamadi, A. Chandra, and B. Singh, “Real-Time HIL
Implementation of Sliding Mode Control for Standalone System Based
on PV Array Without Using Dumpload,” IEEE Trans. Sustainable
Energy, vol. 6, no. 4, pp. 1389-1398, Oct. 2015.
[44] A. Elrayyah, F. Cingoz, and Y. Sozer, “Smart Loads Management Using
Droop-Based Control in Integrated Microgrid Systems,” IEEE Journ. of
Emerging & Selected Topics in Power Electron., vol.5, no.3, pp.1142-
1153, 2017.
[45] R. R. Chilipi, B. Singh, and S. S. Murthy, “Performance of a Self-
Excited Induction Generator With DSTATCOM-DTC Drive-Based
Voltage and Frequency Controller, ” IEEE Trans. Energy Conversion,
vol. 29, no3, pp. 545-557, Sept.2014.
[46] T. Hirose and H. Matsuo, “Standalone Hybrid Wind-Solar Power
Generation System Applying Dump Power Control without Dump
Load,”IEEE Trans. Ind. Elect., vol.59, no.2, pp.988-997, Feb.2012.
[47] J. Leuchter, P. Bauer, V. Rerucha, and V. Hajek, “Dynamic Behavior
Modeling and Verification of Advanced Electrical-Generator Set
Concept,” IEEE Trans Ind. Elect., vol. 56, no1, pp. 266-279, Jan. 2009.
[48] G. Dileep and S. N. Singh, “Maximum power point tracking of solar
photovoltaic system using modified perturbation and observation
method,” Ren. and Sust. Energy Reviews, vol. 50, pp. 109-129, 2015.
[49] B. Singh, A. Chandra, and K. Al-Haddad, Power Quality Problems and
Mitigation Techniques, Chichester, U.K.: Wiley, 2015.

1949-3053 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.