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

Comprehensive Control for Microgrid Autonomous

Operation With Demand Response
Amira Nisar and Mini Shaji Thomas, Senior Member, IEEE

AbstractSeveral situations in a microgrid (MG) cause unbal- while moving from interconnected to islanded mode. The
ances leading to voltage and frequency variations. Conventionally strong motivation towards demand response introduces new
MG power balance is ensured by storage battery units along challenges for smart microgrid operation. Demand response
with unscheduled load shedding. This paper presents control
and management strategies to coordinate MG resources includ- has proven to be a means for reducing peak hour stresses
ing generation, smart electric vehicles, and loads for frequency of distribution network [3], [4]. Many recent researches on
and voltage regulation to achieve optimal autonomous operation. DR provide different algorithms for residential DR which
MG control is designed in two layers of network, i.e., neigh- minimizes household electricity bill and user dissatisfaction.
bour area network (NAN) and house area network (HAN). MG A convex programming DR optimization is proposed in [5]
central controller operates at NAN layer, local load controllers
operate at HAN layer, and smart electric vehicle controllers oper- for automatic load management in a smart home. In [6]
ate at both NAN and HAN layers. Two strategiescomprehensive cost minimization problem is solved based on optimal stop-
control (CC) and vehicle aided CCare developed and imple- ping approach for residential load scheduling to participate
mented in 11/0.4 kV urban residential four-feeder distribution in DR. Optimal DR policies can be framed by integrating
system operating as MG. Both strategies enable calculative DR studies with different electricity pricing models (real-
load manipulation for demand response (DR) considering cus-
tomer load priority, thus maximizing customer satisfaction. The time pricing or day-ahead pricing models) [7], [8]. An insight
simulation results validate the proposed strategies providing fre- into the residential DR potential is presented in [9] wherein
quency regulation and DR for MG autonomous operation. The a home energy management (HEM) algorithm is developed.
proposed control and management strategy also provides an on- HEM software tool with advanced GUI controls the household
line tool for household maximum demand allocation and load power consumption.
management.
However most of the previous work on DR is done with
Index TermsMicrogrid central controller (MGCC), smart respect to the power distribution network. DR operation seen
electric vehicle, home area network, adaptive hill climbing, from MG perspective can provide potential flexibility in its
vehicle to grid mode, household demand limit.
control and management. Hence integrating DR in MG opera-
tion introduces a new paradigm to its concept and applicability.
Issues regarding dynamic studies and economic dispatch of
I. I NTRODUCTION distributed generators (DGs) for sustainable MG operation
ICROGRIDS are independent flexible cells operating
M at low voltage distribution network with distributed
resource, mainly renewable resources, responsive and non-
are reported in [10] and [11]. Studies which focus on MG
control and coordination, report different strategies for differ-
ently designed MG networks. Two level coordination control
responsive loads, storage devices and smart electric vehicles. is designed in [12] for hybrid MG operation. Energy manage-
Smart MG operation includes two-way communication, cus- ment system (EMS) determines the operation mode of different
tomer participation and electric vehicle integration both in grid converters. Back-to-back converter tracks MPPT of wind tur-
to vehicle and vehicle to grid mode. Microgrid is required to bine generator using direct torque control (DTC) scheme
have the ability to operate interconnected with the main grid and perturbation and observation algorithm is used for PV
or isolated (which may be intentional or un-intentional), but generator. Two control strategies, namely Single master and
in both the conditions microgrid must operate autonomously multi master strategies are addressed in [13] for MG islanded
stable [1], [2]. There is a need to formulate techniques to operation. Voltage source inverter is used to provide a refer-
manage and control smart MG operation for fast frequency ence for voltage and frequency. Primary as well as secondary
and voltage regulation in case of internal disturbances or frequency regulation is possible here in the islanded MG.
Frequency responsive load control for a microgrid integrat-
Manuscript received June 26, 2015; revised July 7, 2015 and September 23,
2015; accepted December 23, 2015. Paper no. TSG-00733-2015. ing synchronous machine is presented in [14]. Decision tree
A. Nisar is with the Department of Electrical Engineering, approach is used to determine the operation mode of the con-
Jamia Millia Islamia (Central University), New Delhi 110025, India (e-mail: troller. Central DR strategy provides the load control however
[email protected]).
M. S. Thomas is with the Department of Electrical Engineering and the it does not take into consideration the customer choice and
Centre for Innovation and Entrepreneurship, Jamia Millia Islamia (Central preferences of loads. Also in the absence of synchronous
University), New Delhi 110025, India (e-mail: [email protected]). machine, MG frequency regulation will depend upon the con-
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org. trollable resources and storage battery response to control
Digital Object Identifier 10.1109/TSG.2016.2514483 signals. Selective load control strategy for emergency DR with
1949-3053 c 2016 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

Fig. 2. MG Control Strategy.

Fig. 1. MG Basic Architecture.

distributed storage capacity is proposed in [15] for MG fre-

quency regulation. Load control in this work is based on the
load controller frequency measurement. Such a MG control
system will have large number of frequency measuring units
which will add to the cost of the system and accuracy of Fig. 3. Microsource Controller.
such large number of measuring units will also be a major
issue. Electric vehicles take greater share in personal mar- II. MG A RCHITECTURE AND C ONTROLLERS
ket and smart MG which accommodates EV may experience F UNCTIONALITY
increased stress condition. Lopes et al. [16] suggest adaptive The proposed MG comprises of a LV network with domestic
active power control by implementing phase-locked loop at loads, some of which are critical loads, the rest are non-critical
electric vehicle grid interface. It also suggests that a large and can be controlled, controllable MS (SSMT), non control-
fleet of EV can be integrated in power system by adopting lable MS (PV generator & micro wind generator) and a small
centralized EV charging control strategies. However with the size storage unit. There is also a fleet of smart electric vehi-
advanced sensing and control technologies, EV penetration can cles in the MG, which according to the energy imbalance
be more flexible and SEV charging with creative DR programs and their own SOC act as MG resource. MG is centrally
can result in optimal operation of distribution networks. controlled and managed by the MGCC placed at the distri-
This paper presents comprehensive control of responsive bution substation. Load controllers located at the load centers
loads (PRL ) at HAN for smart microgrid autonomous oper- and SEV controllers, provide decentralized control. The basic
ation. A MG central controller is designed to operate at NAN architecture of proposed MG is presented in Fig. 1. Control
for comprehensive control wherein calculative load manip- and management of the MG is supported by a communication
ulation is initiated on the occurrence of MG disturbance, infrastructure which enables exchange of information between
in accordance with available generation, for frequency and MG entities. Control strategy proposed in this paper pro-
voltage restoration. Local load controllers present in HAN vides frequency responsive control and management for MG
connect/disconnect the loads after receiving maximum demand autonomous operation after unintentional islanding. The com-
signal, taking into account the priority and preferences set prehensive control is provided with MGCC and LCs. The CC
by the house owner. The control flexibility of SEV is com- when combined with grid integration of SEVs provides vehicle
bined with CC to develop VACC, which complements the aided comprehensive control for more effective and optimum
MG autonomous operation and provides fast resilience from MG operation. As shown in Fig. 2, the control strategies have
disturbances by enabling smart charging. been designed in 2-network layers, NAN and HAN layers.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

NISAR AND THOMAS: CC FOR MG AUTONOMOUS OPERATION WITH DR 3

Fig. 5. SBS Control Strategy.

Fig. 4. AHC Control Strategy.

Following subsections explain the functions and operation of

the controllers employed for the MG autonomous operation.

A. Central Controller Functionality

The MGCC operates at NAN layer with the frequency mea-
sured at the point of common coupling (PCC) of the MG
as the input variable. It operates on the frequency deviation
to balance the generation and demand in real time. MGCC
has full control over the reference point settings of the con- Fig. 6. Local Load Controller.
trollable distributed generators. Whenever there is frequency
deviation MGCC sends the control signal to controllable MS
to change the Po (active power reference point) set point to and evaluate as well as control the load demand of household
a new pre-defined value as given in Fig. 3, where according to the received demand limit signal. The local load
fmeasured = Operating frequency of the MG controller as shown in Fig. 6, bears a load control priority-list,
fref = Reference frequency set for the MG operation pre-set and uploaded by the house owner.
f= frequency deviation
Pold
o = SSMT active power set point at the time of 
n 
l
frequency deviation Here, pRLij (t) = pij (t) (1)
Pnew
o = SSMT new active power set point after frequency i j=1
deviation.
However this change in Po is always associated with gener- MDi =Demand limit (KW) assigned to house i
ator constraints and cannot fully satisfy the power balance Pdesirednew,i =Desired demand of ith house after DR
situation of the MG. For complete control of frequency, com- PTi,t =Total power consumption of ith house at time
prehensive control of responsive loads is initiated through instant, t
adaptive hill climbing (AHC) and step by step (SBS) control Pij,t =Power consumption of jth appliance of ith house at
strategy [14]. Under this control scheme maximum demand time instant, t
limit to be assigned to each house is determined by calculat- PRLi,t =Total responsive load of ith house at tth instant
ing the percentage of responsive loads to be disconnected or PNRLi,t =Total non responsive load of ith house at tth instant
reconnected in the MG. Fig. 4 & Fig. 5 illustrates the adaptive i=House participating in demand response
hill climbing & step by step control operations. j=Non-critical/responsive appliance of the house
l=Total number of non-critical/responsive appliance of the
house.
B. MG Local Load Controller Functionality The loads are controlled according to the priority list to min-
MG local load controller operates at HAN layer. It is char- imize customer discomfort. It is actually at this layer the
acterized to monitor the operating state of the household loads demand response actions take place until the total demand
 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

whenever the frequency deviation from the reference fre-

quency is more than the frequency dead band. In case of
less power generation or high power demand in the MG,
frequency of the MG drops below the lower limit of the fre-
quency dead band for SEV controller, the SEVs with SOC
more than 80% moves into V2G mode and controller ini-
tiate power injection (discharging) action. As soon as more
vehicles come in grid integration while participating in power
injection the overall power generation of MG increases and
the frequency regulation is aided. This improves the MG con-
trol and management and provides better utilization of MG
resources.

III. OVERALL MG C ONTROL AND

M ANAGEMENT S TRATEGY
MG generation and load demand imbalance cause frequency
or voltage deviations and affect the stability and autonomous
operation. MG control and management is achieved by adopt-
ing one or more of the following schemes:
Integrating SEV for electric power injection or consump-
tion
Fig. 7. Smart Electric Vehicle Controller with p-f droop. PSEV =Active Initiating demand response actions for responsive resi-
power capacity of smart electric vehicle (kW). Kp = Slope of frequency dential loads.
droop (KW/Hz). Pmin = Maximum discharging power limit of the SEV.
Pmax = Maximum charging power limit of the SEV. Varying power output of the controllable power
resources of MG.
MG control and management shown in Fig. 8 provides
adequate framework to fully control and manage the MG
of the house stays under the maximum demand limit allo-
operation. Three controllers which are integrated to form the
cated to the house by the MGCC. The LLC checks in time
control loops of MG are microgrid central controller, local load
steps that the household demand does not exceed the assigned
controller and smart electric vehicle controller. Disturbances
demand limit by denying the demand request from control-
in MV network, sudden changes in the MG non-controllable
lable smart appliances in accordance with household pre-set
resources (due to sudden change in wind speed and irradiance
preference. The household loads are categorized into critical
level) or sudden change in the amount of load connected to
loads and non-critical loads. Non-critical loads are controllable
the MG result in MG unbalance depicted by P in Fig. 8 and
loads and form the priority list of responsive loads according
power balance equation below in (2).
to the customer comforts. These controllable loads are the DR
action participants at HAN layer, the calculative operation of p = pSSMT + ppv + pWT pSEV pgrid pL (2)
which brings MG frequency into desired limits for MG steady
state operation. Where,
PSSMT = Power output of SSMTs
PPV = Power output of solar generator
C. MG Vehicle Controller Functionality PWT = Power output of micro wind generator
MG vehicle controller is integrated in smart electric vehi- Pgrid = Power import or export from the main grid
cles where it is designed to operate at both NAN and HAN PL = Total load demand.
layer because vehicle load is mobile and may come in Energy imbalance of the MG, for a specified period of time
grid integration outside the house area also. VC operates in is denoted by E. TdP is the time delay due to response of
dual mode system, i.e., V2G mode (discharging) and G2V voltage source inverter and Tinv is the time delay of SEV grid
mode (charging). The charging/discharging control is per- interface inverter. P gives rise to frequency deviation f
formed through p-f droop control strategy as shown in Fig. 7. and MGCC under such situations sends control signals to MS
The implementation of this approach requires phase- locked controllers to change their power output. If the disturbance
loop to measure frequency deviations at each EV grid inter- still persists, on the basis of frequency deviation and respon-
face. The EV active power set point of inverter interface sive loads connected in the MG, maximum demand for each
is adjusted using frequency droop loop. The dead-band and house is calculated and sent to the network of Local load
slope of the frequency droop may be decided by the vehicle controllers. Local load controllers initiate demand response
owners willingness to participate in MG frequency regula- according to the house owners load priority list and con-
tion or it may be decided by the MG operator depending nect/disconnect the responsive loads so that the total demand
upon the MG system requirement. The controller initiates of the house remains within maximum limits. This control
charging (G2V mode) or discharging (V2G mode) actions functionality is explained in Section II and is termed as
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

NISAR AND THOMAS: CC FOR MG AUTONOMOUS OPERATION WITH DR 5

Fig. 8. MG overall management and control.

Fig. 9. MG Test System Model.

comprehensive control of the MG. Total maximum demand bands of CC and VACC is important for this type of
corresponds to Pdesired for MG steady state operation. There MG management.
may be a situation in the MG when sufficient amount of
responsive loads for manipulating (connecting/disconnecting)
are not available (e.g., sudden excessive increase or decrease IV. MG T EST S YSTEM
in wind speed or irradiance level) to achieve steady state. For The microgrid test system used in the study is based on
this kind of situation emergency dispatch from SEV is ini- 500kVA, 11/.433kV LV distribution network [17] as given
tiated. MG control system integrates SEV with controllers below in Fig. 9. Four distributed generators, namely a solar
which sense the system frequency and provide frequency photovoltaic (PV) generator, a micro-wind turbine, and two
regulation by injecting/consuming electric power. This pro- single shaft micro turbines are connected at different feeder
vides enhanced control of the MG emergency situation and nodes to be able to operate the system as autonomous MG. The
is termed as vehicle aided comprehensive control. This type detailed modeling and simulation of distributed resources are
of MG control ensures better utilisation of MG resources available in [18][20]. The overall microgrid generation and
by exploiting electric vehicle flexibility and household load loads are summarized in Table I&II. A storage battery unit of
controllability effectively. Proper setting of frequency dead 30kVA is added in the test network to prevent frequency going
 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

TABLE I TABLE III

MG G ENERATION MG H OUSEHOLD R ESPONSIVE L OADS AND C USTOMER P RIORITY &
P REFERENCES

TABLE II
M ICROGRID L OADS

Fig. 10. Power Generation of MG for islanding at 4.8secs.

out of range in no-control case. The loads are categorised

as responsive loads and non-responsive loads. The responsive
household loads which include the domestic appliances like
hot water heater, air conditioning units, electric ovens, washing
machines, water pumps and clothes dryer are the candidates of
load priority list as presented in Table III. To verify the effec-
tiveness of demand response strategy, loads are modified into
Fig. 11. Frequency, Voltage and Load connected for No control,
constant impedance loads, dependent on frequency and volt- Comprehensive control and Vehicle aided comprehensive control.
age. The microgrid loads are also comprised of electric vehicle
connected to the network throughout single phase chargers
providing fast charging mode (approx. 5-8 hours to reach com-
V. S IMULATION R ESULTS
plete autonomy). Smart electric vehicle is modeled as battery
with 3.3kW charge power. The frequency dead band of micro- A. Scenario 1: Unintentional Islanding
grid central controller is taken as 0.05Hz and that of electric In this case microgrid was importing 184KW from the
vehicle controller is taken as 0.1Hz. In order to match the real- main grid while a sudden fault occurred at MV network and
time situations, 50 vehicles are modeled with state of charge thus MG is disconnected from the main grid. MGCC fixes
more than 80% and rest with different state of charge less the reference power of SSMTs at their nominal values which
than 80%. prior to islanding were producing half of their nominal power.
The MG is assumed to have an electric vehicle fleet Fig. 10 shows the MG generation before and after islanding.
of 96 SEVs willing to participate in vehicle to grid mode Frequency, voltage and the amount of load restored in this
for power injection and primary frequency regulation. In case is presented in Fig. 11. Frequency begins to drop after
order to validate proposed microgrid control strategy com- the fault is initiated and islanding takes place at 3.5 seconds.
plete simulation model is developed in MATLAB Simulink In no-control case it is brought back to normal by unscheduled
environment. load shedding by disconnecting one feeder. In comprehensive
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

NISAR AND THOMAS: CC FOR MG AUTONOMOUS OPERATION WITH DR 7

Fig. 12. Frequency, Voltage at PCC and Active load connected in the MG
with sudden increase in wind speed. Fig. 13. Frequency, voltage at PCC and load connected in the MG for sudden
increase in load demand.

control case MGCC sense the frequency deviation and respon-

sive loads are curtailed. Here maximum demand allocated to loads are reconnected at HAN. The vehicle controller while
each house begins to reduce in time steps and responsive loads monitoring frequency, initiates smart charging of the EVs
are manipulated (shed) by LC at HAN, until the frequency is when the frequency exceeds 60.1 Hz and in this SEV enters
stabilized at 7.8 secs. Under vehicle aided comprehensive con- into G2V mode. The EV with SOC less than 100% partici-
trol VC enters into V2G mode when the frequency falls below pate in MG frequency regulation in this scenario. Frequency
59.89Hz and SEVs starts injecting power to the MG until the regulation is achieved at around 6.9 secs and 7.5 secs and
frequency deviation is zero at 6.1 secs. In this scenario it is the MG active loads after reconnection are 416 and 440kW
observed that 28 SEVs out of the fleet of 96 vehicles suffi- with VACC and CC respectively. In no-control case a stor-
ciently aided frequency regulation. In VACC MG frequency age battery of 30kVA is connected for charging to restore
as well as voltage reaches desired value at a faster rate with MG stability. As shown in Fig. 12, frequency and voltage regu-
higher amount of restored loads. Figure 11 also shows the lation are fastest in the VACC with less amount of manipulated
operation of SBS controller to lower the amount of discon- loads as compared to CC case. In this scenario SBS controller
nected load at around 17.3secs and 15.2secs for CC and VACC operation can be seen at 15.9secs and 13.1secs for VACC and
respectively. It can be seen that the load restored after VACC, CC respectively.
CC and no-control is 490kW, 415kW and 375kW respectively.
C. Scenario 3: Sudden increase in load demand
B. Scenario 2: Sudden increase in uncontrollable In this case there is a sudden change in load demand inside
MS generation the MG resulting in negative frequency deviation. The com-
There may be a situation in the MG when there is a sudden prehensive controller sense this and the responsive loads are
increase in wind speed and hence increased power generation shed. Frequency is stabilized at 6.7 secs. and 7.3 secs. for
by WT. This scenario is simulated and results are reported VACC and CC respectively to achieve MG steady state. SBS
in Fig. 12 with no control, CC and VACC. At 3secs wind control is initiated at 15.1secs for VACC and at 17.4 secs
speed starts increasing and WT generation reaches 190kW for CC. In VACC SEVs with SOC more than 80% move
from 116kW. At 6.1secs. SSMT generation is reduced from into V2G mode and MG frequency as well as voltage reg-
100kW to 50kW with MS controller receiving signal from the ulation is achieved quickly. Fig. 13 illustrates the operation of
MGCC. In comprehensive control, MGCC sense the positive no control, comprehensive control and vehicle aided compre-
frequency deviation and the responsive load control is initiated hensive control for such a situation. It clearly indicates that
when measured frequency exceeds 60.05Hz limit. Responsive almost 100% loads are restored in VACC and 65% loads are
 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

TABLE IV
C OMPARISON OF N O -C ONTROL , CC AND VACC to check the effectiveness of proposed control strategy and the
results (not presented in the paper) showed similar trends.

VI. C ONCLUSION
This paper presents the advanced control and management
strategy for smart microgrid operation. A comprehensive con-
trol of responsive loads is designed in two layer network to
enable demand response considering customer comfort and
load priorities. Microgrid scenarios under islanding and post
islanding (sudden change in MG generation/load) are evalu-
ated through dynamic simulation to demonstrate the effective-
ness of the proposed control and management strategy. The
comprehensive control (CC) provides a solution for MG fre-
quency and voltage regulation by calculative manipulation of
TABLE V responsive loads in case of variations in renewable power gen-
S UMMARY OF THE R ESULTS FOR THE T WO S TRATEGIES eration or load demand with limited storage capacity. It also
eliminates the need of unscheduled load shedding of feeders
as in no-control case. Comprehensive control complemented
with smart electric vehicles enable vehicle aided comprehen-
sive control (VACC), where the MG resilience is faster and
less amount of loads are manipulated. Thus VACC further
expand the controllability of MG to manage even extreme
emergent situations. The proposed strategy also provides an
on-line tool for household maximum demand allocation and
load management. This functionality of the tool sends maxi-
mum demand limit to each house within the microgrid. It also
enables energy management within house area network while
maintaining consumer preferences.

R EFERENCES
[1] B. Lasseter, Microgrids [distributed power generation], in Proc. IEEE
Power Eng. Soc. Winter Meeting, vol. 1. Columbus, OH, USA, 2001,
pp. 146149.
[2] R. H. Lasseter and P. Piagi, Extended microgrid using (DER) dis-
tributed energy resources, in Proc. IEEE PES Gen. Meeting, Tampa,
FL, USA, Jun. 2007, pp. 15.
[3] S. Shao, M. Pipattanasomporn, and S. Rahman, Grid integration of elec-
restored in CC. In no-control case two feeders are disconnected tric vehicles and demand response with customer choice, IEEE Trans.
to restore MG stability. As shown in Fig. 13 SBS control is Smart Grid, vol. 3, no. 1, pp. 543550, Mar. 2012.
[4] M. S. Thomas and A. Nisar, Study of demand response in Indian resi-
initiated at 17.3 secs for VACC and at 15.1 secs. for CC. dential sector, in Proc. Nat. Conf. Emerg. Trends Elect. Electron. Eng.,
The simulation results reveal that MG frequency and voltage vol. 1. 2015, pp. 283291.
regulation is better with CC as compared to no-control case. [5] K. M. Tsui and S. C. Chan, Demand response optimization for smart
home scheduling under real-time pricing, IEEE Trans. Smart Grid,
The amount of loads restored, as shown in Table IV, are higher vol. 3, no. 4, pp. 18121821, Dec. 2012.
with CC because in no-control case complete feeders are [6] P. Yi, X. Dong, A. Iwayemi, C. Zhou, and S. Li, Real-time opportunistic
disconnected for steady state operation which jeopardizes cus- scheduling for residential demand response, IEEE Trans. Smart Grid,
vol. 4, no. 1, pp. 227234, Mar. 2013.
tomer QoS and is thus undesirable. With VACC MG resilience [7] H. Saele and O. S. Grande, Demand response from household cus-
is much faster and the loads restored increase further. Table V tomers: Experiences from a pilot study in Norway, IEEE Trans. Smart
shows that maximum demand allocated to each customer is Grid, vol. 2, no. 1, pp. 102109, Mar. 2011.
[8] S. Li, D. Zhang, A. B. Roget, and Z. ONeill, Integrating home energy
higher in VACC for insufficient MG generation situation. In simulation and dynamic electricity price for demand response study,
different scenarios of MG autonomous operation with VACC, IEEE Trans. Smart Grid, vol. 5, no. 2, pp. 779788, Mar. 2014.
it is seen that the number of electric vehicles involved in charg- [9] M. Pipattanasomporn, M. Kuzlu, and S. Rahman, An algorithm for
intelligent home energy management and demand response analysis,
ing or discharging is 28 and 49 out of a fleet of 96 vehicles, IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 21662173, Dec. 2012.
while eliminating the need of external storage batteries. This [10] S.-J. Ahn, S.-R. Nam, J.-H. Choi, and S.-I. Moon, Power scheduling of
provides and optimal solution for MG autonomous operation. distributed generators for economic and stable operation of a microgrid,
IEEE Trans. Smart Grid, vol. 4, no. 1, pp. 398406, Mar. 2013.
Environmental factors leading to unexpected changes in [11] S. Bae and A. Kwasinski, Dynamic modeling and operation strategy for
uncontrollable MS output and sudden unpredictable changes in a microgrid with wind and photovoltaic resources, IEEE Trans. Smart
load cause MG transient situations. Therefore other scenarios Grid, vol. 3, no. 4, pp. 18671876, Dec. 2012.
[12] X. Liu, P. Wang, and P. C. Loh, A hybrid AC/DC microgrid and its coor-
with sudden change in MS generation (like change in irradi- dination control, IEEE Trans. Smart Grid, vol. 2, no. 2, pp. 278286,
ance level) and load demand (decrease) were also simulated Jun. 2001.
 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.

NISAR AND THOMAS: CC FOR MG AUTONOMOUS OPERATION WITH DR 9

[13] J. A. P. Lopes, C. L. Moreira, and A. G. Madureira, Defining control Amira Nisar received the B.Tech. degree in electri-
strategies for microgrids islanded operation, IEEE Trans. Power Syst., cal engineering and the M.Tech. degree in electrical
vol. 21, no. 2, pp. 916924, May 2006. power system management from the Department
[14] S. A. Pourmousavi and M. H. Nehrir, Real-time central demand of Electrical Engineering, Jamia Millia Islamia,
response for primary frequency regulation in microgrids, IEEE Trans. New Delhi, India, where she is currently pur-
Smart Grid, vol. 3, no. 4, pp. 19831996, Dec. 2012. suing the Ph.D. degree. She is a Lecturer of
[15] C. Gouveia, J. Moreira, C. L. Moreira, and J. A. P. Lopes, Coordinating Electrical Engineering with the G. B. Pant Institute
storage and demand response for microgrid emergency operation, IEEE of Technology, Directorate of Training and Technical
Trans. Smart Grid, vol. 4, no. 4, pp. 18981908, Dec. 2013. Education, New Delhi. Her research interests include
[16] J. A. P. Lopes, F. J. Soares, and P. M. R. Almeida, Integration of smart grid, microgrid, renewable energy, demand
electric vehicles in the electric power system, Proc. IEEE, vol. 99, response, and electric vehicles.
no. 1, pp. 168183, Jan. 2011.
[17] S. Ingram, S. Probert, and K. Jackson. (Oct. 2003). The
Mini Shaji Thomas received the graduation degree
Impact of Small Scale Embedded Generation on the Operating
from the University of Kerala, the M.Tech. degree
Parameters of Distribution Networks, PB Power. [Online]. Available:
from IIT Madras, both with gold medals, and
http://webarchive.nationalarchives.gov.uk/20100919181607
the Ph.D. degree from IIT Delhi, India. She is
[18] M. S. Thomas and A. Nisar, Data-driven modeling and simulation of
a Professor with the Department of Electrical
PV array, in Proc. IEEE Int. Conf. Comput. Sustain. Glob. Develop.
Engineering, Jamia Millia Islamia (JMI), where she
(INDIACom), New Delhi, India, 2015, pp. 308313.
is the Director of the Centre for Innovation and
[19] MATLAB/Simulink. (Mar. 5, 2013). SimPowerSystems Documentation.
Entrepreneurship. She has set up the first of its kind
[Online]. Available: http://www.mathworks.com/
Supervisory Control and Data Acquisition (SCADA)
[20] Y. Zhu and K. Tomsovic, Development of models for analyzing the
Laboratory and Substation Automation Laboratory
load-following performance of microturbines and fuel cells, Elect.
at JMI, and started a unique M.Tech. Program
Power Syst. Res., vol. 62, no. 1, pp. 111, May 2002.
in Electrical Power System Management. She is the Coordinator of the
[21] H. Wu, M. Shahidehpour, and A. Al-Abdulwahab, Hourly demand
Special Assistance Program on Power System Automation. She has authored
response in day-ahead scheduling for managing the variability of renew-
a textbook entitled Power System SCADA and Smart Grids, and published
able energy, IET Gener. Transm. Distrib., vol. 7, no. 3, pp. 226234,
around 120 research papers in international journals and conferences of
Mar. 2013.
repute. Her research areas include SCADA systems, substation and distri-
bution automation, and smart grid. She is a Distinguished Lecturer of the
IEEE Power and Energy Society.