This article has been accepted for inclusion in a future issue of this journal.

Content is final as presented, with the exception of pagination.

IEEE TRANSACTIONS ON SMART GRID 1

A New Framework for Microgrid Management:

Virtual Droop Control
Ashishkumar Solanki, Member, IEEE, Adel Nasiri, Vijay Bhavaraju, Yakov L. Familiant, Member, IEEE,
and Qiang Fu, Member, IEEE

Abstract—Microgrids can provide the most promising means Microgrids can be reconnected to the utility grid without any
of integrating large amounts of distributed sources into the power interruption once the utility is recovered. They can provide
grid and can supply reliable power to critical loads. A generalized higher reliability, energy security and surety, and open the
control framework is required to regulate microgrid voltage and
frequency, maintain system stability, maintain power quality, and door to significant system efficiency improvements [3]. It also
manage distributed generations. A microgrid control framework opens the door for electricity market players in terms of ancil-
for power management and voltage and frequency regulation lary services. Consumers and businesses can supply valuable
is proposed in this paper. The proposed method, virtual droop services including demand response, real-time price response,
control, is described and formulated and compared with the exist- and voltage support to the grid in return for payments from
ing natural droop control technique. Unit commitment algorithm
has also been implemented to manage nonrenewable sources to the serving utility or independent system operator [3].
improve the system efficiency. The proposed technique operates Managing DGs (renewables, conventional, and storage)
the microgrid at a constant voltage and frequency, and uses com- and loads within microgrid during island and grid-tie modes
munication for power sharing. It also provides the means to and transitions is a challenge. Some of the key micro-
operate the microgrid in case of lost communication and sabo- grid control scheme concerns are: 1) maintaining stability;
tage on communication network. The proposed method has been
applied to Fort Sill microgrid and the modeling results have 2) regulating voltage and frequency; 3) proper active and
been compared with natural droop control technique. A labora- reactive load sharing; and 4) easier and faster islanding and
tory setup, which consists of a 100 kW natural gas generator, reconnection [4]–[8]. Due to the limited availability of capac-
56 kWh Li-ion based battery with 250 kW inverter, and 100 kW ity overhead of the sources, the stability of a microgrid is of
load bank, has been built and tested. The results of the setup concern, especially during the starting of motor loads, large
have been provided, which confirms the viability of the proposed
technique. loads and interface with wind and solar photovoltaic (PV)
that are highly intermittent. The microgrid needs to be stable
Index Terms—Distributed generation (DG), droop control, throughout these changes to the demand and the capacities
microgrid, power management, unit commitment.
of the energy storage and renewable sources. The microgrid
controller needs a method of commanding the sources and
a method of assessing the available capacity of the sources
I. I NTRODUCTION within its command.
ICROGRID is a group of interconnected loads and There are several control methods proposed previously
M distributed generation (DG) within clearly defined elec-
trical boundaries that acts as a single controllable entity with
for power management and to control voltage and fre-
quency within microgrid in island mode. The Consortium for
respect to the grid [1]. Microgrid can operate in parallel to the Electric Reliability Technology Solutions (CERTS) microgrid
grid or as an island. The most compelling feature of a micro- concept proposed autonomous controller in which, voltage
grid is the ability to separate and isolate itself from the utility’s and frequency stability are achieved by drooping the volt-
distribution system unintentionally during events (i.e., faults, age and frequency according to active and reactive power
voltage collapses, and blackouts) [2], [3]. It may also inten- requirement [4], [5], [9]. The natural droop method is not
tionally disconnect during grid maintenance and also when the suitable when the microgrid has nonlinear loads due to the
quality of power from the grid falls below certain standards. harmonic current. Moreover, all the resources in the micro-
grid contribute power to the load and operate autonomously;
Manuscript received December 20, 2013; revised July 23, 2014, opportunities to optimize the system and improve efficien-
January 24, 2015, and May 29, 2015; accepted July 5, 2015. cies are limited. Virtual impedance droop control method
Paper no. TSG-00931-2013.
A. Solanki and A. Nasiri are with the University of Wisconsin–Milwaukee, was proposed to modify the output impedance for better
Milwaukee, WI 53201 USA (e-mail: [email protected]). power sharing [10]. But the output impedance has to be
V. Bhavaraju and Q. Fu are with Eaton Corporation Innovation Center, designed and it will add one more variable to control the
Menomonee Falls, WI 53051 USA (e-mail: [email protected];
[email protected]). design. An angle droop control method was proposed in
Y. L. Familiant is with Nazarbayev University, Astana 010000, Kazakhstan which voltage angle is drooping instead of frequency for
(e-mail: [email protected]). power sharing for less frequency variation [11]. However,
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. angle measurement with common-reference is needed and
Digital Object Identifier 10.1109/TSG.2015.2474264 fast communication is required with angle droop control.
1949-3053 c 2015 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 inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

2 IEEE TRANSACTIONS ON SMART GRID

A modified droop control method is proposed in [12], which

uses an adaptive feedforward compensation for stability
enhancement during transient but preserves the conventional
droop mechanism during steady state [12]. In a hierarchical
control method, discussed in [13] and [14], a secondary con-
trol loop is added to change the reference power for each
source to restore the frequency to 1 p.u. Voltage restoration
is based on introducing a virtual impedance for inverter based
sources. This method is based on natural droop control as
are many other methods. The power sharing is still based Fig. 1. Active and reactive power dynamics of the natural gas generator.
on modified droop control. Although the hierarchical control
shows good performance, it takes seconds (more than 1 s) for and interface of intermittent power, accurate modeling of nat-
each loading/unloading event to restore frequency. For a sys- ural gas generator and energy storage inverter is essential.
tem with variable load or generation profile (in presence of Therefore, brief modeling and testing of natural gas generator
renewables), this performance can pose a serious challenge. and energy storage inverter are discussed below.
In all the natural droop based methods, voltage and frequency
needs to vary in order to apply power sharing between sources. A. Natural Gas Generator
The main objective of this paper is to introduce a new frame-
work for microgrid control, power management, and reducing Natural gas generator plays a very important role in
voltage and frequency variations. It also offers improved effi- the microgrid. The exciter is in charge of reactive power
ciency, better utilization of renewable energy, and minimizes and governor adjusts the active power. The exciter AC8B
the usage of natural gas generator by using a unit commit- has been modeled according to IEEE Standard 421.5 [17].
ment algorithm. Secure communication is used to improve The automatic voltage regulator in this model consists of
efficiency, power quality, and power management. The pro- a proportional-integral-derivative control, with separate con-
posed control will provide the capability to operate a microgrid stants for proportional, integral, and derivative gains. The
in case of lost communication or any sabotage on the commu- values of constants are chosen for the best performance of
nication network. It provides flexibility to integrate more DGs each particular generator excitation system. The developed
in the future and the proposed framework can be applied to model for generator includes time delays in amplifier, exciter
other microgrids as well. Detailed comparison between pro- dynamics, and field time constant as well as gas engine
posed virtual droop control (VDC) and natural droop control dynamics.
has been performed in this paper. The microgrid system stud- To verify the developed model, several load tests have been
ied in this paper is modeled based on the real system data and performed and the results from model have been compared
is simulated using PSCAD software. Experimental results of with the experimental results. In standalone mode, step loads
a test setup have also been presented in this paper to provide have been applied to verify the dynamic behavior of the nat-
further insights to the proposed technique. ural gas generator. Isochronous control and synchronization
scheme have also been implemented for the two synchronous
generators. Fig. 1 shows the active and reactive power dynam-
II. M ICROGRID C ONFIGURATION AND ics when the isochronous control for the generator is applied.
C OMPONENT M ODELING At the 13th s, 100% power command with a power factor of
The proposed controller has been applied to Fort Sill micro- 0.8 is applied. At the 30th s, power command changes from
grid with ratings of 480 V, 60 Hz, and 630 kW. This microgrid 100% to 10%.
is connected to the utility grid through a 480 V/13.20 kV
transformer and a static switch. The generations include two B. Voltage Source Converter System
natural gas generators each rated at 190 kW, one 90 kW solar The renewable energy and energy storage systems uti-
PV system, a 2.5 kW wind turbine, and a 250 kW energy lize voltage source converters (VSCs) to interface with
storage device. grid/microgrid. There are mainly two modes of operation, cur-
The solar PV and wind turbine generators are con- rent mode and voltage mode, for VSC. A current mode VSC
nected to the system through inverters operating in a current is generally used to extract maximum power from renew-
mode [5]–[7] and the energy storage inverter is operated in able energy resources. A voltage mode control is used for
a voltage mode [15]. The system also includes various motor the energy storage system in the proposed control technique
loads and variable loads. Motor loads mainly include chillers, is this paper.
water pumps, and air compressors. During grid-tie operation, 1) Current Mode VSC System: In the current mode VSC
natural gas generators are turned off. In an island mode of system, the line current is tightly regulated by a dedicated cur-
operation, natural gas generators are brought into the sys- rent control scheme, through the converter side ac voltage. The
tem according to developed unit commitment algorithm. The real and reactive powers are controlled by the angle and mag-
detailed modeling of the solar PV and wind generators are nitude of the line current with respect to the terminal voltage.
provided in [16]. To ensure the stability of the proposed con- The terminal voltage and line current are first transformed to
trol technique during load change, transition between modes a d-q reference frame and the regulation is performed in the
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 3

system frequency and voltage remain constant. By tightly reg-

ulating voltage and frequency, the reconnection mechanism to
grid, from an island mode, becomes easier and faster.
The energy storage inverter is operating in a voltage mode
and natural gas generators are operated in active/reactive
power control mode. However, wind and PV inverters are
operated in a current mode. This means that energy storage
inverter provides a voltage with a magnitude of 1 p.u. and
frequency of 60 Hz. Other sources, including natural gas gen-
erators and renewables must synchronize with this voltage to
deliver power. All the sources in a microgrid communicate
Fig. 2. Schematic of the current mode VSC system. with each other using TCP/IP Modbus secure communication
protocol [21], [22].

A. Unit Commitment Algorithm

A unit commitment strategy is applied for the natural
gas generators to increase the efficiency and reduce the fuel
consumption. For various levels of loads and the state of
charge (SOC) of the energy storage, none, one, or both
generators are synchronized to supply power.
A unit commitment strategy is applied for the natural gas
generators to increase the efficiency. For various levels of loads
and the SOC of the energy storage, none, one, or both gener-
ators are synchronized to supply power. For instance, during
Fig. 3. Schematic of the voltage mode VSC system.
light load condition with a higher percentage of SOC, none
of the generators are connected and only energy storage pro-
vides power. Details of the unit commitment procedure for
rotating frame. Then the reference values are transferred to three modes are discussed below.
the stationary reference frame and applied to the converter. Mode 1: Only energy storage provides power to the system.
Fig. 2 shows the schematic of the current source VSC sys- Mode 2: Energy storage and natural gas generator 1 provide
tem used for renewables in the proposed microgrid control of power to the system.
this paper. Mode 3: Energy storage and both natural gas generators
The control parameters for the 90 kW solar PV system are: provide power to the system.
Kp1 = 0.03, Ki1 = 0.5, Kp2 = 0.007, and Ki2 = 0.1. Fig. 4 shows the algorithm of the unit commitment. The
Simulation tests have been conducted on this model and total load is measured by adding the energy storage power to
the results have been verified with the experimental system the power of generators
testing.
2) Voltage Mode VSC System: In the voltage mode VSC PLoad = PL + Ploss − Ppv − Pwind = PNG1 + PNG2 + PES .
system, the amplitude and the angle of the ac output voltage (1)
is regulated and consequently the active and reactive power
of the device is controlled when operating in parallel with If the system load is less than PL1_H kW and the SOC of
other sources. The voltage mode VSC system can form a grid energy storage is greater than SOCmin %, then the energy stor-
or provide power to loads when the grid is disconnected. age supplies the total microgrid load and natural gas generators
Fig. 3 shows the control schematic for the Fort Sill micro- are off to improve efficiency. In this mode, if the system load
grid energy storage inverter [18]. The control parameters of exceeds PL1_H kW, natural gas generator 1 is turned on to
the voltage mode VSC system are as follows: Kp1 = 0.1, provide power. However, if the system load is less than PL1_H
Ki1 = 10, Kp2 = 0.4, and Ki2 = 0.2. kW but the SOC of energy storage is less than SOCmin %, then
in order to charge the energy storage, the natural gas gener-
ator 1 is again connected, and the system moves to mode 2.
The system is moved to mode 1 if the load is decreased to
III. VDC F RAMEWORK AND U NIT less than PL1_L kW and SOC is increased to above SOCmax %.
C OMMITMENT A LGORITHM A hysteresis band PL1_L −PL1_H is defined to prevent frequent
The natural droop control method has been proposed in the generator turn on and off. While in mode 2, if the system load
literature to regulate the voltage and frequency in a micro- is increased to above PL2_H kW, generator 2 will turn on and
grid in an island mode [4]–[8], [19], [20]. The voltage and moves the system to mode 3. In addition, when the total load
frequency are drooped naturally to contribute active and reac- is less than PL2_L kW but the SOC is less than SOCmin %, then
tive power to the load. In the proposed VDC framework, the the system still moves from modes 2 to 3. In mode 3, when
power sharing is performed between various sources while the the system load is less than PL2_L kW and the SOC is greater
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

4 IEEE TRANSACTIONS ON SMART GRID

Fig. 5. Generalized virtual frequency and voltage droop curves.

to system voltage. A virtual voltage is determined according

to reactive power output of the energy storage inverter. The
virtual voltage will determine the reactive power command
for natural gas generators from a droop relationship, defined
between the virtual voltage of the system and reactive power
command of each source. It should be noted that since energy
storage inverter is placed in a voltage mode, it supplies the
difference between load active and reactive power and other
sources in the microgrid. It behaves as a slack bus in a power
system concept. Power commands of natural gas generators
are updated only when load variation is greater than defined
value. Load variation less than defined value is taken care of
by the energy storage inverter.
Virtual frequency and voltage slopes can be defined for
different modes by user for different microgrid architecture.
Fig. 5 shows the slope curve for mode n and m generator.
The output active and reactive power of the energy stor-
age is measured and the virtual frequency and voltage are
Fig. 4. Unit commitment algorithm as part of VDC to increase efficiency. determined from the measured value. The active and reac-
tive power commands for the generators are determined from
than SOCmax %, the generator 2 is turned off and the system virtual frequency and voltage.
is moved to mode 2. The parameters for unit commitment are For any mode n, PES max , PES min , PES_atratedF , QES max ,
selected based on several criteria. QES min , QES_atratedV , Fmax , Fmin , Frated , Vmax , Vmin ,
1) The power output of natural gas generators should be Vrated , PNG max , PNG min , and PNG_atrated are defined by
close to their rated power. the user. These parameters could be different for different
2) To increase the life expectancy of the energy storage, it mode and different generator. For different mode, the virtual
is necessary to operate it in the linear operating region frequency and virtual voltage have been calculated based on
of open circuit voltage-SOC. the defined curves as
3) During island mode of operation, the SOC of the energy
For PES max < PES_measured < PES_atratedF
storage should be greater than 50% at all the time.  
The size of energy storage and generators also plays an impor- Frated − Fmax
Frated =
tant role in the unit commitment parameter selection. Based on PES max − PES_atratedF
 
the above criteria, the parameters selected for unit commitment × PES_measured − PES_atratedF + Fmax
algorithm for Fort Sill microgrid are as follows: PL1_L = 120 For PES_atratedF < PES_measured < PES_ min
kW, PL1_H = 170 kW, PL2_L = 270 kW, PL2_H = 310 kW,  
Fmin − Frated
SOCmin = 60%, and SOCmax = 90%. Frated =
PES_atratedF − PES_ min
 
B. Virtual Droop Control × PES_measured − PES_ min + Frated (2)
To regulate the active and reactive power output of the For QES_ max < QES_measured < QES_atratedV
 
sources, a virtual frequency and voltage are created. The active Vrated − Vmax
Vrated =
power output of the energy storage inverter determines the vir- QES_ max − QES_atratedV
tual frequency from virtual droop curve. The droop curve is  
× QES_measured − QES_atratedV + Vmax
defined between energy storage active power output and vir-
For QES_atratedV < QES_measured < QES_ min
tual frequency. The virtual frequency will determine the active  
power commands for natural gas generators from a droop Vmin − Vrated
Vrated =
relationship, defined between the virtual frequency and active QES_atratedV − QES_ min
 
power command of each source. The same concept applies × QES_measured − QES_ min + Vrated . (3)
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 5

Fig. 6. Control flow diagram of microgrid controller in island mode.

Based on the virtual frequency and voltage, the power

command is calculated as

For Fmax < Fvirtual < Frated

 
PNG_atratedF − PNG max Fig. 7. Virtual frequency and voltage droop curves for (a) mode 2,
PNG_CMD = when energy storage and one generator provide power in island mode, and
Fmax − Frated (b) mode 3, when energy storage and both generators provide power in
× (Fvirtual − Frated ) + PNG max island mode.
For Frated < Fvirtual < Fmin
 
PNG min − PNG_atratedF
PNG_CMD =
Fmax − Frated
× (Fvirtual − Frated ) + PNG_ratedF (4)
For Vmax < Vvirtual < Vrated
 
QNG_atratedV − QNG max
QNG_CMD =
Vmax − Vrated
× (Vvirtual − Vrated ) + QNG max
For Vrated < Vvirtual < Vmin
 
QNG min − QNG_atratedV
QNG_CMD = Fig. 8. Active load profile for the microgrid.
Vmax − Vrated
× (Vvirtual − Vrated ) + QNG_ratedV . (5)

Fig. 6 shows the control flow diagram of microgrid con- IV. C OMPARISONS B ETWEEN VDC AND
troller during island mode and when communication presents. NATURAL D ROOP C ONTROL
Fig. 7 shows virtual droop curves for modes 2 and 3 for Fig. 8 shows the active and reactive load profile of the
Fort Sill microgrid studied. As shown in Fig. 7(a), for instance, microgrid used for analysis. In the system modeling, a 24-h
when microgrid load is at 230 kW, the virtual frequency profile is scaled down to 240 s duration. Same load profile
is 59.75. From the virtual frequency versus active power com- is used to implement the natural droop control method for
mand curve, the 170.2 kW power command is sent to natural comparison. The droop curves for natural droop control are
gas generator. Once, the natural gas generator sets its power provided in Fig. 9. The energy storage and generator volt-
output to 170.2 kW, energy storage power is automatically age and frequency are adjusted according to their active and
adjusted to 59.8 kW. reactive power output.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

6 IEEE TRANSACTIONS ON SMART GRID

Fig. 9. Frequency and voltage droop curves for natural droop method. (a)

(b)

Fig. 11. Voltages and frequency when natural droop control method is
applied (a) voltage at PCC terminal, energy storage device, natural gas
generators 1 and 2, and (b) microgrid frequency.

Fig. 10. Active and reactive power when natural droop control method
is applied; active power at PCC terminal, energy storage, natural gas
generators 1 and 2 (top), reactive power for the same devices (bottom).

Initially, microgrid operates in the grid-tie mode. It is

assumed that in this mode, the energy storage is fully charged.
The system moves from grid-tie mode to island at 35 s. The
operation of the microgrid in grid-tie mode is the same for both
methods. In case of natural droop control method, as soon
as the point of common coupling (PCC) breaker is opened,
signals are sent to both generators to turn them on.
As shown in Fig. 10, natural gas generators are synchro-
nized at the 49th s and droop control is activated. After the
49th s, both natural gas generators and energy storage inverter Fig. 12. Active power (top) and reactive power (bottom) at PCC terminal,
adjust the voltage and frequency based on their output power energy storage device, natural gas generators 1 and 2, reactive power; when
and predefined curves. Since the droop curves and rating of VDC method is applied.
both generators are the same, they share equal amount of active
and reactive power. The maximum active and reactive power
output from both generators are 133.5 kW and 97.43 kVAR, decides the turn on–off operation of the natural gas gener-
respectively, which is quite less than their power rating. They ators. As the microgrid load is greater than 170 kW at the
are operating at lower efficiency compared with their nominal 38.4th s (Fig. 8), natural gas generator 1 is turned on. As
efficiency. shown in Fig. 12, at the 48th s, natural gas generator 1 is syn-
As shown in Fig. 11(a), the voltages vary between 0.456 and chronized and the system operates in case 2. At the 89.25th s,
0.504 kV based on the system reactive load. Fig. 11(b) shows microgrid load becomes greater than 310 kW and natural gas
that the frequency varies between 59.5 and 60.5 Hz based on generator 2 is turned on and synchronized at the 102.9th s,
the system active load. and the microgrid enters case 3 as shown in Fig. 12. At the
However, in case of VDC method, as soon as the PCC 199.6th s, the microgrid load drops below 270 kW, which turns
breaker opens, unit commitment algorithm is activated and off the natural gas generator 2. At the 223rd s, the microgrid
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 7

Fig. 13. Voltage at PCC terminal, energy storage device, natural gas gen-
erators 1 and 2 (top), and microgrid frequency (bottom); when VDC control
method is applied.

load drops below 120 kW and natural gas generator 1 is turned

off as well and system moves to case 1.
Natural gas generators are operated very close to the rated
active and reactive power. The maximum active and reac-
tive power of the natural gas generator 1 are 188.8 kW and
133.3 kVAR, respectively, while for natural gas generator 2,
the output power is at 181.2 kW and 126 kVAR. This method
of operation improves the efficiency of the natural gas gen- (a) (b)
erators. Fig. 13 shows the system voltage and frequency for
Fig. 14. Control flow diagram of “control transfer process” at (a) energy
this case. The voltages at DGs and energy storage remain very storage inverter and (b) natural gas generators.
close to 0.480 kV. The small voltage differences are required to
transfer power. Microgrid frequency remains at 60 Hz during
island mode for VDC method, as shown in Fig. 13(b). From the last information on all the sources, each genera-
There are some spikes on the frequency and voltage when tor decides their ID based on predefined criteria. The same
a generator is connected/disconnected or a step load is applied. criteria are used for all the generators to avoid any conflict
The durations of these spikes are very small. Some spikes between the generators’ ID numbers. Each DGs and energy
are attributed to how frequency is calculated. Any change or storage (ES) inverter have real-time clock, which provides
transient in the voltage waveform can appear as a short-term a tick signal at every 2 s. In communication lost event, based
frequency change. on a tick signal the system control transfer process can be ini-
tialized. Before starting the transfer process, each source, hold
till the next tick signal to avoid any undesired situation. All the
V. O PERATION W ITHOUT C OMMUNICATION sources check frequency and voltage and if they are at rated
Even though the proposed control methodology uses secure values, the generator with ID number 1 becomes the master
communication, it is necessary to study the microgrid in case and it will switch its control from active/reactive power control
of lost communication. The proposed control framework pro- mode to voltage/frequency control. At the same tick signal, the
vides the means to transfer the control from VDC to natural energy storage inverter will switch to current mode from volt-
droop and operate without the communication. Energy stor- age mode. The other generator remains in the active/reactive
age inverter plays very important role in this control transfer. power control mode. Based on the current generator ID and
Fig. 14 shows the control flow diagram of control transfer the corresponding last power output of that generator, energy
process. storage will move its power up or down with a predefined
During normal operation, microgrid operates according to slope. As the power from the energy storage inverter changes,
VDC control method and unit commitment algorithm. Once the output power of master generator will also change because
communication loss is detected, each DG including the energy the master generator is in voltage/frequency control mode and
storage, operate according to the last image of the system. moves toward the natural droop operating point. Once the
For instance, natural gas generator 1 has the last power out- master generator reaches this point, it will switch its con-
put information of generator 2 and energy storage inverter. trol from voltage/frequency control to droop control. Since the
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

8 IEEE TRANSACTIONS ON SMART GRID

Fig. 15. Transfer controls from VDC to natural droop control.

Fig. 16. Microgrid frequency (top), active power at PCC, energy storage,
other generator and energy storage inverter does not know the natural gas generators 1 and 2 (bottom); during the transition from VDC to
status of the master generator, energy storage inverter keeps natural droop.
moving up or down. Since the master generator is already in
droop mode, the frequency and voltage will change. When fre-
quency changes F Hz from nominal frequency of 60 Hz, the
second generator and energy storage will come to know that
master generator is in droop mode. Energy storage inverter
moves its power in reverse direction and brings the frequency
back to 60 Hz and wait for the next tick signal. At the next
tick signal, the generator with ID number 2 becomes mas-
ter and switches its control from active/reactive power control
to voltage/frequency control. Energy storage inverter repeats
the same procedure to move the second generator to its droop
curve operating point. Once all the generators switch their con-
Fig. 17. Schematic of the experimental test setup.
trols to droop control, energy storage slowly moves its power
up/down and finds its own droop curve. The changes in load
during control transfer process can delay the process. However, moves its control from active/reactive power control to volt-
it will not impact the operation of the microgrid, since at least age/frequency control. At point D, the second generator also
one source will be in voltage mode during the entire process. reaches at the droop curve operating point and switches its
An example of the operation can be described when consid- control to droop control. At point E, all other sources come
ering that the microgrid load is 400 kW and the communica- to know that generator 2 is operating in the droop mode.
tion is lost between the sources. As shown in Fig. 15, the last Then energy storage inverter changes its power output and
image is that the generator 1 provides 190 kW, generator 2 pro- at point F it switches its control to droop control. After all
vides 180 kW, and energy storage provides 30 kW. Since both the sources switch to droop control, the microgrid can operate
generators’ powers are above the natural droop curves, energy autonomously without communication.
storage inverter has to increase its power output in order to
bring the generators to the natural droop curves. Natural gas
generator 1 is assigned ID 1 and generator 2 is assigned ID 2. VI. E XPERIMENTAL S ETUP FOR THE VDC C ONCEPT
Fig. 16 shows the frequency and active powers of energy In order to experimentally verify the techniques developed
storage and generators during control transfer process. At in this paper, an experimental test setup has been built. The
point A, generator 1 switches its control from active/reactive schematic of the setup is shown in Fig. 17 and pictures of
power control to voltage/frequency control and energy storage component used for the experiment are shown in Fig. 18.
switches to current mode. Energy storage inverter increases its A SEL 849 relay has been used to measure the microgrid
power output to bring the generator 1 to the droop curve. At load and send data to storage inverter using TCP/IP Modbus
point B, generator 1 reaches at the droop curve and switches secure communication protocol. WT1600 power meter is used
its control to natural droop control. At point C, frequency to measure power, voltage, and frequency. A 100 kW natu-
changes by 0.2 Hz and the second generator and energy stor- ral gas generator and a 250 kW storage inverter are used for
age come to realize that generator 1 is switched to droop the experimental setup. The rating of each unit of Altairnano
control. At point C, the second generator becomes master and power battery is 125 kW and one of the units is used at a time.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 9

Fig. 20. Microgrid frequency (top), microgrid voltage (bottom); experiment

results.

Fig. 18. (a) SEL 849 rely to measure total load and send data to micro-
grid inverter using Modbus TCP/IP. (b) WT1600 power meter to measure the
powers, frequency and voltages. (c) 100 kW load bank. (d) Altairnano energy
storage. (e) SMAX inverter connected to energy storage. (f) 100 kW natural
gas generator.

Fig. 21. Active power at PCC, natural gas generators, energy storage and
solar PV; deployments of VDC control method at the Fort Sill microgrid
testbed.

synchronized and can provide power. Active power of the

generators are initially limited at 90 kW and then raised to
100 kW. After the 250th s, a reconnection command is sent to
the system. Since VDC regulates the system voltage and fre-
quency at 1 p.u., the reconnection process is fast with minimal
transients.
Fig. 19. Active power at natural gas generator, energy storage, and microgrid
load; experiment results.
VII. S TABILITY A NALYSIS
In traditional large power systems, frequency can be main-
Once the generator is synchronized with the microgrid tained by large inertia of synchronous generators and active
inverter, a 35 kW power command is given to the generator. power regulation, while the voltage is maintained by adjusting
Any change in the load is taken care of by the energy stor- reactive power. The major sources for reactive power are syn-
age inverter since the inverter is operating in a voltage mode. chronous generators, synchronous condensers, and capacitor
Fig. 19 shows the load and active power of natural gas gener- banks. Regulating transformers also contribute to regulating
ator and energy storage. Fig. 20 shows the microgrid voltage system voltage. In a microgrid with high-renewable pene-
and frequency when VDC is applied. Larger variations in fre- tration, large numbers of inverter-based sources and energy
quency compared with the modeling results occur due to the storage provide power to the loads. Hence, during an island
existence of a low-pass filter on the slew rate of the energy mode of operation, small inertia support is available to main-
storage inverter. tain the frequency. Therefore, the stability analysis of any
Fig. 21 shows the behavior of the system and control con- control technique for a microgrid concept is required. A state
cept when it is applied to the 650 kW Fort Sill microgrid. space model has been developed for the testbed Fort Sill
Initially the microgrid operates in a grid-tie mode when unin- microgrid utilizing VDC method in order to perform a stability
tentional islanding occurs. Energy storage device picks up the analysis. The analyses include several states for each of the
majority of the load before natural gas generators 1 and 2 are components, energy storage inverter, natural gas generators,
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

10 IEEE TRANSACTIONS ON SMART GRID

Fig. 24. Poles and zeros map for mode 2, when energy storage, renewables,
and one generator provide power in island mode.

Fig. 22. Step response from vdref to ves_od (top) and from vqref to ves_oq
(bottom).

Fig. 25. Poles and zeros map for mode 3, when energy storage, renewables,
and both generators provide power in island mode.

fast dynamics based on the locations of the poles, confirmed

by the simulation and experimental results.
A natural gas generator is turned on when the system moves
to mode 2 (Fig. 24). Several slow poles are added to the sys-
tem, which can be attributed to the slower dynamics of exciter
and governor. However, the system stability is maintained. For
mode 3 (Fig. 25), when natural gas generator 2 and more dis-
Fig. 23. Poles and zeros map for mode 1, when only energy storage and
renewables provide power in island mode. tribution line are also connected, poles further move to right
but the system stability is still preserved.
renewable inverters, and distribution lines. Load active/reactive
power are considered as input to the state space. The important VIII. C ONCLUSION
states and the detail state space equations of each components
In this paper, a VDC method is proposed and introduced to
are given in the Appendix.
manage microgrids in various operating modes. The proposed
Initially, individual state space equations were developed
method has been applied to a 650 kW testbed microgrid. It
for each component with 10 states for ES, 19 states for each
regulates the microgrid voltage and frequency more tightly
natural gas, 8 states for each renewable inverter, and 4 states
than the natural droop based control techniques. It also oper-
for each distribution line. Real parameters for each component
ates the fossil based generators at near rated power to increase
have been used to develop the equations. Then, a final state
their operating efficiency. The technique to operate microgrid
space model of the system with dimension of 92 was devel-
without communication is also presented. A state space model
oped by integrating the individual sets. The developed model
of the entire testbed microgrid integrating the proposed con-
utilizes the VDC method by placing the energy storage device
trol concept has been developed and stability of the system
in a voltage mode and natural gas generators in PQ mode.
has been analyzed by mapping the poles-zeros. The simu-
Renewable sources are placed are operated in a current mode.
lation results are presented for the Fort Sill microgrid and
The developed model enables detailed stability analyses of the
compared with the results from basic natural droop control
proposed method.
method. Experimental results of a test system in the labora-
The step response from vdref to ves_od and from vqref to
tory as well as Fort Sill microgrid have been presented when
ves_oq for the energy storage inverter are shown in Fig. 22.
applying the proposed technique.
The device has a very fast dynamic response as expected to
support any load change.
State space poles and zeros placement for the testbed micro- A PPENDIX
grid is depicted in Figs. 23–25. Fig. 23 shows the pole-zero The important states of the state space model are as follows.
map of the system in mode 1, when only energy storage ies_od , ies_oq : d-axis and q-axis ES output current.
inverter and renewables are connected. The system has very ves_od , ves_oq : d-axis and q-axis ES output voltage.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 11

ipv_od , ipv_oq : d-axis and q-axis PV output current. B. State Space Modeling of Voltage Mode Inverter
iwind_od , iwind_oq : d-axis and q-axis wind output current. The voltage mode VSC system can form a grid or provide
ωng1 , Vng1 : Generator 1 speed and terminal power to loads when the grid is disconnected. The close loop
voltage. state space equations of voltage mode inverter can be written
ωng2 , Vng2 : Generator 2 speed and terminal as follows:
voltage.  
d  Res1 + Kesp1d Kesp1d ∗ Kesp2d
vdref_es , vqref_es : d-axis and q-axis ES reference voltage. ies_id = − ies_id − ves_cd
idref_pv , iqref_pv : d-axis and q-axis PV reference current. dt Les1 Les1
Kesp1d ∗ ω0 (3 ∗ Ces )
idref_wind , iqref_wind : d-axis and q-axis wind reference − ves_cq
current. Les1
Pref_ng1 , Qref_ng1 : Generator 1 active and reactive power Kesi1d Kesp1d ∗ Kesid2
+ Xesd1_int + Xesd2_int
reference. Les1 Les1
Pref_ng2 , Qref_ng2 : Generator 2 active and reactive power Kesp1d Kesp1d ∗ Kesp2d
+ ies_od + vdref (14)
reference. L Les1
 es1 
d  Res1 + Kesp1q Kesp1q ∗ Kesp2q
ies_iq = − ies_iq − ves_cq
dt Les1 Les1
A. State Space Modeling of Current Mode Inverter Kesp1q ∗ ω0 (3 ∗ Ces )
The control of the current mode inverter system is based on + ves_cd
Les1
the d-q reference frame. Hence, the mathematical equations of Kesi1q Kesp1q ∗ Kesiq2
the LCL filter are derived in d-q reference frame. The current + Xesq1_int + Xesq2_int
Les1 Les1
controller is also represented in a mathematical equation to Kesp1q Kesp1q ∗ Kesp2q
close the loop, which gives the input voltage vren_id and vren_iq . + ies_oq + vqref (15)
Les1 Les1
The close loop state space equations of current mode inverter
d  1  
can be written as follows: ves_cd = ω0 ves_cq + ies_id
dt 3C
   es
1 Ns  
d  Rren1 + Krenpd Krenpd − ies_od (16)
iren_id = − iren_id + idref 3Ces Np
dt Lren1 Lren1
Krenid d  1  
+ Xrendq_int (6) ves_cq = −ω0 ves_cd + ies_iq
dt 3C
L
 ren1    es
1 Ns  
d  Rren1 + Krenpq Krenpq − ies_oq (17)
iren_iq = − iren_iq + iqref 3Ces Np
dt Lren1 Lren1
Kreniq d 
+ Xrenq_int (7) Xesd1_int = Kesid2 Xesd2_int + Kespd2 vdref − Kespd2 ves_cd
Lren1 dt
− ω0 (3 ∗ Ces )ves_cq + ies_od − ies_id (18)
d 
Xrendq_int = idref − iren_id (8) d 
dt Xesq1_int = Kesiq2 Xesq2_int + Kespq2 vqref − Kespq2 ves_cq
d  dt
Xrenq_int = iqref − iren_iq (9) + ω0 (3 ∗ Ces )ves_cq + ies_oq − ies_iq (19)
dt
d  1   d 
vren_cd = ω0 vren_cq + iren_id Xesd2_int = vdref − ves_cd (20)
dt 3C dt
 ren d 
1 Ns   Xesq2_int = vqref − ves_cq (21)
− iren_od (10) dt
3Cren Np  
d  Res2 1 Np
d  1   ies_od = ω0 ies_oq − ies_od + ves_cd
vren_cq = −ω0 vren_cd + iren_iq dt Les2 Les2 Ns
dt 3C
  ren  
Ns   1 Np 2
−
1
iren_q (11) − ves_od (22)
Les2 Ns
3Cren Np  
d  d  Res2 1 Np
iren_od = ω0 iren_oq −
Rren2
iren_od ies_oq = −ω0 ies_od − ies_oq + ves_cq
dt Lren2 dt Les2 Les2 Ns
   2  
1 Np 1 Np 1 Np 2
+ vren_cd − vren_od − ves_oq . (23)
Lren2 Ns Lren2 Ns Les2 Ns
(12) C. State Space Modeling of Natural Gas Generator
d  Rren2
iren_oq = −ω0 iren_od − iren_oq When natural gas generator is synchronized and connected
dt Lren2
   2 to the microgrid, the stability of the microgrid is crucial due
1 Np 1 Np to different time responses of generators and energy storage
+ vren_cq − vren_oq .
Lren2 Ns Lren2 Ns system. The detail state space models of the exciter and natural
(13) gas engine are developed with sufficient details.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

12 IEEE TRANSACTIONS ON SMART GRID

d
The complete state space equations for small perturbation (xa3 ) = 0 (39)
which includes all the mechanical dynamics are as follows: dt
d KA KIR KA (KPR + KIR TDR )
d  (VR ) = xa1 + xa2
XPng_int = Png_ref − Eng_od ing_od − Ing_od eng_od dt TA TA
dt KA (KPR TDR + KDR ) 1
− Eng_oq ing_oq − eng_oq Ing_oq (24) + xa3 − VR
TA TA
d (40)
(xc1 ) = xc2 (25)
dt d KE 1
d  (EFD ) = VR − EFD (41)
XPng_int = Png_ref − Eng_od ing_od − Ing_od eng_od dt TE TE
dt d 1 1
− Eng_oq ing_oq − eng_oq Ing_oq (26) (Vc ) = −  Vc +  EFD . (42)
dt Td0 Td0
d
(xc1 ) = xc2 (27) A distribution transmission line in a studied microgrid can
dt
d 1 b1 1 1 be represented by parameters in a proportional-integral section
(xc2 ) = − xc1 − xc2 + ωref − ω̄r and the state space equations of distribution line are given as
dt b2 b2 b2 b2
Ki_Png Kp_Png follows:
+ XPng_int + Png_ref d 2 2
b2 b2 (vind ) = ω0 vinq + (iind ) − (iiTd ) (43)
Kp_Png dt CiT CiT
− Eng_od ing_od d  2   2  
b2 vinq = −ω0 vind + iinq − iiTq (44)
Kp_Png dt CiT CiT
− Ing_od eng_od d RiT 1 1
b2 (iiTd ) = ω0 iiTq − iiTd − voutd + vind (45)
Kp_Png dt LiT LiT LiT
− Eng_oq ing_oq d  RiT 1 1
b2 iiTq = −ω0 iiTd − iiTq − voutq + vinq (46)
Kp_Png dt LiT LiT LiT
− eng_oq Ing_oq (28) d 2 2
b2 (voutd ) = ω0 voutq + (iTd ) − (ioutd ) (47)
d  1 a1 1 dt CiT CiT
xg1 = xc1 + xc2 − xg1 (29) d  2   2  
dt d1 d1 d1 voutq = −ω0 voutd + iTq − ioutq . (48)
d  dt CiT CiT
xg2 = 0 (30)
dt R EFERENCES
d G1 G1 c1 1
(u4 ) = xg1 + xg2 − u4 (31) [1] DOE Microgrid Workshop Report, U.S. Dept. Energy, San Diego, CA,
dt e1 e1 e1 USA, Aug. 2011.
d [2] P. Asmus, “Why microgrids are inevitable and why smart utilities should
(u5 ) = u4 (32) plan accordingly,” Distrib. Energy Mag., pp. 44–47, Sep./Oct. 2011.
dt [3] R. L. Dohn. (2011). “The business case for microgrids,” White
d  1 1 Paper Siemens. [Online]. Available: http://w3.usa.siemens.com/
T̄m = − T̄m + u5 (33) smartgrid/us/en/microgrid/Documents/The%20business%20case%20for
dt Tch Tch
%20microgrids_Siemens%20white%20paper.pdf, accessed Jan. 19,
dω̄r 1 Eng_od
= T̄m − ing_od 2014.
dt 2H 2H [4] R. H. Lasseter and P. Paigi, “Microgrid: A conceptual solution,” in Proc.
Ing_od Eng_oq IEEE 35th Annu. Power Electron. Special Conf. (PESC), vol. 6. Aachen,
− eng_od − ing_oq Germany, 2004, pp. 4285–4290.
2H 2H [5] P. Piagi and R. H. Lasseter, “Autonomous control of microgrids,” in
Ing_oq Proc. Power Eng. Soc. Gen. Meeting, Montreal, QC, Canada, 2006,
− eng_oq (34) pp. 1–8.
2H [6] N. Hatziargyriou, H. Asano, R. Iravani, and C. Marnay, “Microgrids,”
dδ
= ωo ω̄r . (35) IEEE Power Energy Mag., vol. 5, no. 4, pp. 78–94, Jul./Aug. 2007.
dt [7] F. Katiraei, R. Iravani, N. Hatziargyriou, and A. Dimeas, “Microgrids
management,” IEEE Power Energy Mag., vol. 6, no. 3, pp. 54–65,
The complete state space equations of excitation system for May/Jun. 2008.
small perturbation are as follows: [8] R. H. Lasseter, “Extended CERTS microgrid,” in Proc. IEEE Power
Energy Soc. Gen. Meeting Convers. Del. Elect. Energy 21st Century,
d  Pittsburgh, PA, USA, 2008, pp. 1–5.
XQng_int = Qng_ref − Eng_od ing_oq − Ing_oq eng_od [9] J. Eto et al., “Overview of the CERTS microgrid laboratory test bed,” in
dt Proc. CIGRE/IEEE PES Joint Symp. Integr. Wide-Scale Renew. Resour.
− Eng_oq ing_od − eng_oq Ing_od (36) Power Del. Syst., Calgary, AB, Canada, 2009, p. 1.
d [10] J. Kim, J. M. Guerrero, P. Rodriguez, R. Teodorescu, and K. Nam,
(xa1 ) = xa2 (37) “Mode adaptive droop control with virtual output impedances for an
dt inverter-based flexible AC microgrid,” IEEE Trans. Power Electron.,
d 1 1 1 vol. 26, no. 3, pp. 689–701, Mar. 2011.
(xa2 ) = − xa2 − Vc + V1ref [11] R. Majumder, A. Ghosh, G. Ledwich, and F. Zare, “Angle droop versus
dt TDR TDR TDR frequency droop in a voltage source converter based autonomous micro-
KQng_i KQng_d grid,” in Proc. Power Energy Soc. Gen. Meeting, Calgary, AB, Canada,
+ XQng_int + Qng_ref 2009, pp. 1–8.
TDR TDR [12] M. B. Delghavi and A. Yazdani, “An adaptive feedforward compensation
KQng_d for stability enhancement in droop-controlled inverter-based microgrids,”
− Qng (38) IEEE Trans. Power Del., vol. 26, no. 3, pp. 1764–1773, Jul. 2011.
TDR
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

SOLANKI et al.: NEW FRAMEWORK FOR MICROGRID MANAGEMENT: VDC 13

[13] Y. A.-R. I. Mohamed and A. A. Radwan, “Hierarchical control system Vijay Bhavaraju received the B.S. degree in electrical engineering from
for robust microgrid operation and seamless mode transfer in active dis- the Indian Institute of Technology Madras, Chennai, India, in 1976;
tribution systems,” IEEE Trans. Smart Grid, vol. 2, no. 2, pp. 352–362, the M.S. degree in power system operation and controls from Sri
Jun. 2011. Venkateswara University, Tirupathi, India, in 1988; and the Ph.D. degree in
[14] C.-X. Dou and B. Liu, “Multi-agent based hierarchical hybrid control for power electronics from Texas A&M University, College Station, TX, USA,
smart microgrid,” IEEE Trans. Smart Grid, vol. 4, no. 2, pp. 771–778, in 1994.
Jun. 2013. He was in the oil industry designing and commissioning offshore and land
[15] V. Bhavaraju, A. Nasiri, and Q. Fu, “Multi-inverter controls and man- rigs. He was with Tech Power Systems, Carrabelle FL, USA (later acquired by
agement of energy storage for microgrid islanding,” Elect. J., vol. 25, NOV), where he developed three important products: the mud-pump synchro-
no. 8, pp. 36–44, Oct. 2012. nizer, the auto-drill, and block controller. He was with Ford-Ecostar, Deroit,
[16] Q. Fu et al., “Microgrid generation capacity design with renewables and MI, USA, from 1998 to 2004, researching inverters for microturbines, photo-
energy storage addressing power quality and surety,” IEEE Trans. Smart voltaics, and fuel cells. Since 2005, he has been with the Corporate Research
Grid, vol. 3, no. 4, pp. 2019–2027, Dec. 2012. and Technology Group, Eaton Corporation Innovation Center, Menomonee
[17] IEEE Recommended Practice for Excitation System Models for Power Falls, WI, USA. He has been involved in different projects related to invert-
System Stability Studies, IEEE Standard 421.5-1992, 1992. ers for solar, batteries, and microgrids. He led a team that released the 250 kW
[18] A. Yazdani and R. Iravani, Voltage-Sourced Converters in Power PV inverter.
Systems: Modeling, Control, and Applications. Hoboken, NJ, USA: Dr. Bhavaraju was a Member of the IEEE 1547 Standard from 2000 to
Wiley, 2010. 2004. He is currently a Member of the IEC Project Team of the Microgrid
[19] A. L. Dimeas and N. D. Hatziargyriou, “Operation of a multiagent sys- for Disaster Preparedness and Recovery.
tem for microgrid control,” IEEE Trans. Power Syst., vol. 20, no. 3,
pp. 1447–1455, Aug. 2005.
[20] J. A. Peas Lopes, C. L. Moreira, and A. G. Madureira, “Defining control
strategies for microgrids islanded operation,” IEEE Trans. Power Syst.,
vol. 21, no. 2, pp. 916–924, May 2006.
[21] A. Swales, Open Modbus/TCP Specification, Schneider Elect.,
Rueil-Malmaison, France, 1999.
[22] Q. Xinxiao and J. Zhiping, “Automation monitoring and control system
based on Modbus/TCP protocol [J],” Comput. Eng., vol. 30, no. 8, p. 69,
2004.

Yakov L. Familiant (S’01–M’05) received the

B.S.E.E. degree from Northwest Technological
Ashishkumar Solanki (S’10–M’14) was born in University, St. Petersburg, Russia, in 1995, and the
Anand, India, in 1985. He received the B.S. degree M.S.E.E. and Ph.D. degrees from the University
from Sardar Patel University, Vallabh Vidhyanagar, of Wisconsin–Milwaukee, Milwaukee, WI, USA, in
India, in 2007; the M.S. degree from Gannon 2001 and 2006, respectively, all in electrical engi-
University, Erie, PA, USA, in 2008; and the neering.
Ph.D. degree from the University of Wisconsin– He was a Postdoctoral Fellow with the University
Milwaukee (UWM), Milwaukee, WI, USA, in 2015, of Wisconsin–Madison, Madison, WI, in 2006, and
all in electrical engineering. the Naval Research Laboratory, Washington, DC,
He was a Research Assistant with UWM, from USA, in 2007. In 2008, he joined Eaton Corporation
2011 to 2015, and a Research and Development Innovation Center, Milwaukee. In 2014, he joined Nazarbayev University,
Intern with Tapco, Milwaukee, in 2010. He was Astana, Kazakhstan, as an Assistant Professor. His current research interests
also with Corporate Research Group, Eaton Corporation Innovation Center, include power electronics, microgrids, and control.
Menomonee Falls, WI, in 2013 and 2014. His current research interests
include microgrids, wind energy, solar photovoltaics, energy storage systems,
and power quality and power management in microgrids.

Adel Nasiri was born in Sari, Iran, in 1974. He

received the B.S. and M.S. degrees from the Sharif
University of Technology, Tehran, Iran, in 1996 and
1998, respectively, and the Ph.D. degree from the
Illinois Institute of Technology, Chicago, IL, USA,
in 2004, all in electrical engineering. Qiang Fu (S’10–M’13) was born in Anhui
He is currently an Excellence in Engineering Province, China, in 1984. He received the B.S.
Faculty Fellow of Power Electronics and the and M.S. degrees in electrical engineering from
Director with the Center for Sustainable Electrical Chongqing University, Chongqing, China, in 2006
Energy Systems, Department of Electrical and 2009, respectively, and the Ph.D. degree
Engineering and Computer Science, University of in electrical engineering from the University
Wisconsin–Milwaukee, Milwaukee, WI, USA. His current research interests of Wisconsin–Milwaukee, Milwaukee, WI, USA,
include renewable energy interface, energy storage, and microgrids. in 2013.
Dr. Nasiri is currently an Editor of the IEEE T RANSACTIONS ON He is currently an Electrical Engineer with the
S MART G RID and Power Components and Systems, and an Associate Corporate Research and Technology Group, Eaton
Editor of the IEEE T RANSACTIONS ON I NDUSTRY A PPLICATIONS and the Corporation Innovation Center, Menomonee Falls,
International Journal of Power Electronics. He was the General Chair of the WI. He has co-authored the book Architecture, Programming, and Interfacing
2012 IEEE International Symposium on Sensorless Control for Electrical for the Freescale DSP 56F8346 (China Machine Press, 1999), published over
Drives, the International Conference on Renewable Energy Research and ten conference and journal papers, and holds one Chinese patent. His current
Applications 2014, and the IEEE Symposium on Power Electronics and research interests include modeling and assessing of microgrids, as well as
Machines for Wind and Water Applications 2014. probabilistic analysis of power systems.