SPECIAL SECTION ON SMART GRIDS: A HUB OF INTERDISCIPLINARY RESEARCH

Received October 1, 2015, accepted October 11, 2015, date of publication November 24, 2015,
date of current version December 10, 2015.
Digital Object Identifier 10.1109/ACCESS.2015.2503379

Demand Response Management for Residential

Smart Grid: From Theory to Practice
WEN-TAI LI1 , (Student Member, IEEE), CHAU YUEN1 , (Senior Member, IEEE),
NAVEED UL HASSAN2 , (Senior Member, IEEE), WAYES TUSHAR1 , (Member, IEEE),
CHAO-KAI WEN3 , (Member, IEEE), KRISTIN L. WOOD1 , KUN HU4 ,
AND XIANG LIU4 , (Member, IEEE)
1 Singapore

University of Technology and Design, Singapore 487372

Engineering Department, Lahore University of Management Sciences, Lahore 54792, Pakistan
of Communications Engineering, National Sun Yat-sen University, Kaohsiung 804, Taiwan
4 School of Software and Microelectronics, Peking University, Beijing 100871, China
2 Electrical
3 Institute

Corresponding author: W.-T. Li (wentai_li@sutd.edu.sg)

This work was supported by the Singapore University of Technology and Design through the Energy Innovation Research Program,
Singapore, under Grant NRF2012EWT-EIRP002-045.

ABSTRACT In recent years, many studies have investigated the potential of demand response management (DRM) schemes to manage energy for residential buildings in a smart grid. However, most of the
existing studies mainly focus on the theoretical design of DRM schemes and do not verify the proposed
schemes through implementation. Smart grid research is highly interdisciplinary. As such, the establishment
of testbeds to conduct DRM requires various skill sets that might not always be possible to arrange. However,
the implementation of a DRM scheme is critical not only to verify the correctness of the design in a
practical environment but also to address many important assumptions that are necessary for the actual
deployment of the scheme. Thus, the theoretical aspect of DRM solutions should be discussed and verified
in a practical environment to ensure that the scheme is suitable for deployment. In this paper, we propose
a DRM scheme and construct a residential smart grid testbed to implement the proposed scheme. In the
proposed DRM scheme, we suggest two different types of customer engagement plans, namely, green
savvy plan and green aware plan, and design algorithms based on two user inconvenience indices to
evaluate DRM for peak load reduction. The testbed verifies the effectiveness and efficiency of the proposed
DRM scheme.
INDEX TERMS Smart grid, user inconvenience, peak load reduction, customer engagement plan, demand
response, energy management service (EMS), implementation.
I. INTRODUCTION

Recently, the demand for electricity has been increasing

as a result of economic and industrial developments. For
instance, a recent annual energy outlook report of the Unite
State Energy Information Administration forecasted that the
residential electricity demand will increase by 24% within
the following several decades [1] because of the substantial
increase in population. A smart grid, with its two-way communication and power flow features, is a potential solution to
meet this increasing demand. In particular, demand response
management (DRM) has been identified as one of the key
components of the smart grid that can help the power market
increase the efficient use of energy by remote monitoring
and control of electricity load and by setting efficient energy
prices with a view to shift the high demand of electricity users
to an off-peak period.

VOLUME 3, 2015

To this end, significant efforts have been exerted in the

past few years to investigate the potential of DRM schemes in
managing energy for residential buildings in the smart grid.
Examples of such studies include recent [2][11] in 2015
and [12][16] in 2014. In [2], an optimization-based home
load control scheme is proposed to manage the operation
periods of responsive electrical appliances. The study also
recommend several operation periods for non-responsive
loads. Two types of customer engagement plan are designed
in [3] to specify the amount of intervention in the customers
load settings to reduce the peak load. Chuan and Ukil [4]
establish a mathematical model to represent and model
the load profile type of residential buildings in Singapore
by describing in detail their energy requirement and consumption patterns. Methodologies for managing the temperature sensitive parts of residential electricity demand, that is,

2169-3536
2015 IEEE. Translations and content mining are permitted for academic research only.
Personal use is also permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

2431

W.-T. Li et al.: DRM for Residential Smart Grid

heating, ventilation, refrigeration, and air conditioning (AC)

are developed in [5] and [6]. In [7] and [8], the authors
investigate energy scheduling techniques with delay-tolerant
demands in residential smart grids. Different techniques
developed in 2015 for residential DRM can be found
in [9][11].
Numerous studies have also been conducted in 2014 to
address the issues of residential DRM. For instance, an energy
consumption scheduling scheme for household appliances is
proposed in [12] to reduce the peak-to-average ratio of the
electricity system. In [13], a distributed direct load control
scheme is proposed with a view to match the actual aggregated demand with the desired aggregated demand profile.
A coordinated DRM scheme is designed in [14] to generate the load profile and minimize the individual cost to the
customers. In [15] and [16], DRM schemes are designed
based on the time-of-use electricity distribution tariff and
customer reward plan, respectively. We stress that numerous
other studies have discussed similar issues, which can be
found in the literature reviewed in [17][20].
Most of the existing studies, mainly focus on the
theoretical design of DRM schemes without any verification through implementation. Smart grid research is highly
interdisciplinary. As such, the establishment of testbeds to
conduct DRM requires various skill sets that might not always
be possible to arrange. However, the implementation of the
DRM scheme is critical not only to verify the correctness
of the design in a practical environment but also to address
many important assumptions that are necessary for actual
deployment of the scheme. For instance, a wireless sensor
network (WSN) is suitable for various control and monitoring applications of the smart grid and has been used
in most smart grid-related studies. Although WSNs can
provide cost-efficient and reliable solutions [21], they are
unsuitable for delay-critical applications, which may be challenged when sudden failures occur in the monitored environment [22]. Thus, the suitability of different communication
facilities should be understood for different smart grid applications [23], which can be addressed once we implement the
system and verify the designed DRM technique.
Several studies have implemented testbeds to validate
potential applications related to the smart grid [24][28].
These testbeds have various aims, scales, limitations, and
features. For instance, in [24], a microgrid testbed is established and combined with a cognitive radio network as
a means of communication. Two research groups have
developed experimental testbeds in laboratory environments
to validate their research as discussed in [25] and [26].
In [27] and [28], the authors demonstrate two microgrid
testbeds in campus environments that contain hybrid energy
sources. However, there has been limited effort in the literature on constructing a testbed for residential DRM studies.
Thus, constructing a testbed for residential DRM will verify
not only the theoretical aspect of DRM solution but also
the schemes that are suitable for deployment in a practical
environment.
2432

To this end, this study proposes a control mechanism for

DRM in residential buildings and implements the scheme in
a smart grid testbed. The main contributions of this study are
described as follows:
In the proposed control mechanism, we suggest
two types of customer engagement plans, namely, Green
Savvy Plan (GSP) and Green Aware Plan (GAP).
We also design algorithms based on two user inconvenience indices to evaluate DRM for peak load
reduction.
For residential DRM, we construct a testbed to implement the proposed scheme and conduct experiments to
verify the effectiveness and efficiency of the proposed
schemes.
The remainder of the paper is organized as follows:
The system model of the proposed scheme is discussed
in Section II. The customer engagement plans are presented in Section III. The control algorithms are developed
in Section IV. The testbed and experimental results are
explained in Section V. The study is concluded in Section VI.

FIGURE 1. Demonstration of the system model considered in this study.

II. SYSTEM DESCRIPTION

The system model considered in this study mainly consists

of three layers based on the types of activities conducted
for DRM and is shown in FIGURE 1. The proposed architecture comprises several households, where each household has numerous electrical appliances. These appliances
are divided into essential and flexible load categories. The
power consumption requirements and scheduling times of
essential loads are fixed and cannot contribute to the DRM.
By contrast, flexible loads can be controlled and their power
consumption and scheduling time can be adjusted according
to the DRM requirements. Lighting load can be an example of
flexible load as we can switch off unnecessary or extra lights
that would result in power saving. AC load can be another
example of flexible as we can switch off or operate the AC at
a higher thermostat set point to save power.
VOLUME 3, 2015

W.-T. Li et al.: DRM for Residential Smart Grid

We adopt an open Internet-of-things architecture, where a

centralized database records the energy usage of all electrical
appliances from all households [29]. Therefore, in the proposed architecture, each household is connected to a home
gateway and several smart appliances that can communicate
with each other. The home gateway is controlled through
a home gateway manager in the cloud. Such architecture
enables multiple third parties to connect to the system to provide services. For example, a home automation and security
company may let households subscribe to its service, where it
helps monitor the home when the owners are away and sound
an alarm to the owners when an intrusion is detected by
various sensor and energy readings in the database. Similarly,
an energy management service (EMS) company can participate in an ancillary energy market, which is the focus of this
case study.
The EMS provider may have a deal with the grid in
providing services to maintain the load below a certain
threshold. Then, the EMS provider recruits households by offering them customer engagement plans that
describe incentive and inconvenience. The objective of the
EMS provider is to earn some profit while meeting the target
load reduction required by the grid operator by recruiting a
sufficient number of residents and offering them appropriate
incentive. We will discuss this concept and implement a
testbed comprising the EMS provider that controls a set of
appliances that represent various households.

TABLE 1. Sample GSP and GAP plans based on time cut and power cut
methods.

in terms of the number of watt-minutes (watt-hours) that

are cut by the EMS provider for DRM and the incentive
is offered as dollar per Kilowatt-hour (KWh) for the power
being cut during the duration. In Table 1, we provide sample
GSP and GAP based on the time cut and power cut methods.1
For example, in the time cut method, GSP users are offered
1.5$/h as incentive, which is 1.5 times compared with that of
the GSP user. As such, GSP users might accept three times
inconvenience at most compared with GAP users and earn
13.5$/day at most when the EMS denied its service 3 h/day to
the GAP user, whereas GAP users accept less inconvenience
(3 h/day at most) but also less incentive of 1$/h, such that
the user in this plan can earn 3$/day at most. By contrast,
in the power cut method, inconvenience is specified in terms
of the total amount of power cut. Similarly, the GSP user
is offered an incentive of 4.5$/kWh and accepts three times
inconvenience at most compared with the GAP user
(i.e., power of the GSP user is cut 3 kWh/day by the EMS
and earn 13.5$ at most). By contrast, for the GAP user, not
more than 1 kWh/day is cut with only 3$/kWh as incentive
(giving the user 3$/day at most).

III. CUSTOMER ENGAGEMENT PLANS

The EMS provider can offer customer engagement plans

to recruit households to participate in DRM to reduce a
threshold, as desired by the grid operator. We propose
customer engagement plans that specify the amount of inconvenience and incentive. In this study, we assume that, for
DRM, the EMS provider divides customers into the following
two categories:
Green savvy: Users that accept more inconvenience but
with more incentive;
Green aware: Users that accept less inconvenience but
with less incentive.
For these customer types, we assume that the EMS provider
defines two different customer engagement plans termed
as (i) GSP and (ii) GAP. These plans specify the inconvenience and incentive for green savvy and green aware users.
We let 0 GSP and 0 GAP denote the inconvenience defined in
the two plans, such that 0 GSP 0 GAP . Similarly, we let
3GSP and 3GAP denote the incentive per unit of inconvenience offered by the EMS provider in the two plans, such
that 3GSP 3GAP .
In this study, we define inconvenience in two ways, namely,
1) time cut and 2) power cut. In the time cut method,
we define inconvenience in terms of number of minutes
(or hours) the user is denied operation of its demanded
appliances (regardless of their power consumption) and the
incentive is offered as dollars per minute (or hour) of inconvenience. In the power cut method, we define inconvenience
VOLUME 3, 2015

IV. PROBLEM FORMULATION AND

CONTROL ALGORITHMS

In this section, we formulate a generic optimization problem,

followed by control algorithms that can be used to implement the customer engagement plans based on inconvenience
indices (time cut or power cut). In this formulation, we associate weights for each user and for each appliance of each
household. Then, these weights are dynamically adjusted
according to the user subscription plans (GSP or GAP) and
the criterion (time cut or power cut) being used by the
EMS provider.
A. PROBLEM FORMULATION

As mentioned in Section II, appliances can be divided into

essential and flexible load categories. Therefore, we let Pej (t)
denote the total essential load of user j at time t. We also
assume that all appliances mentioned in the subsequent
description belong to flexible loads. We let pi,j denote the
rated power consumption of the i-th appliance of user j, and
si,j = {si,j (1), . . . , si,j (T )} denote the demand status vector of
load i of user j, where the time index of DRM goes from t = 1
to t = T and si,j (t) = 1 when appliance i is demanded at
time t by user j and zero otherwise. Thus, di,j (t) = si,j (t)pi,j (t)
denotes the power consumption of appliance i of user j at
1 The exact design of GSP and GAP, that is, the offered values of inconveniences and incentives respectively can be obtained by the EMS provider
using surveys (such that the values are acceptable to the end users).

2433

W.-T. Li et al.: DRM for Residential Smart Grid

time t. Then, the totalPpower demand of user j at time t is

expressed as Pj (t) = i di,j (t) + Pej (t). Then, we obtain the
total power demand at time t, as follows:
X
PU (t) =
Pj (t).
(1)
j

However, the grid operator will require the EMS provider

to keep the total power consumption of the users below a
certain threshold Ppeak , that is, PU (t) Ppeak (t), because of
economic scheduling or stable operation of generators. When
PU (t) > Ppeak (t), the EMS provider will conduct DRM by
switching off several appliances of the users to maintain the
total power demand below the given threshold. Thus, these
users whose appliances are switched off by the EMS provider
will experience inconvenience at time t. As such, the
EMS provider should consider not only the engagement
plans of users but also the inconvenience of users when the
EMS provider conducts DRM.
To this end, we propose a method that associates weight
factors with each user and also with each appliance in each
household. Let j (t) be the user inconvenience weight factor
that controls user js inconvenience based on his subscription
plan and let i,j (t) be the appliance preference factor that
controls the ratio of influence in terms of the i-th appliance
of user j based on his declaration preference. Subsequently,
we formulate an optimization problem whose goal is to minimize user inconvenience and which is subject to total power
demand below the required threshold level Ppeak , and the final
results conform with the subscription plans of users. This
procedure is equivalent to determining the optimal demand
status, such that the sum-weight is minimized at each time
slot. For ease of notation, we drop index (t) from all the
succeeding description because we perform the optimization
problem at each time slot. Thus, the optimization problem is
formulated as
X X
min
j
i,j li,j
li,j , i,j

s.t.

pi,j si,j li,j

D
E+
PU Ppeak

i,j

li,j {0, 1},

i, j,

(2)

where hxi+ = max{x, 0}, and li,j is a binary value. If the

i-th appliance of the j-th user is switched off at this time,
then li,j = 1; otherwise li,j = 0. When the EMS provider
turns off the appliances of certain users at this time slot,
it raises the inconvenience of these users. In this case,
the afected user inconvenience weight factors increase,
thereby reducing the chance of these users being turned off in
the next time slot. Similarly, users can also set their preference
for the appliances to be turned off. In this case, the appliance
preference factor serves a role, such that the less preferred
appliances are turned off first. Using these concepts, the user
inconvenience weight factors and the appliance preference
factors are updated at each time slot.
2434

B. WEIGHT UPDATING

The user inconvenience can be defined based on many

different criteria and presented in many different ways.
As mentioned before, in this paper, we define inconvenience
into two criteria: 1) time cut and 2) power cut. Therefore,
we propose a method that designs a factor called user inconvenience weight factor to indicate and quantify the inconvenience.2 Moreover, we design two user inconvenience weight
factors based on both criteria. Such weight factors are adapted
to the proposed plans and also consider the fairness for users
subscribing to the same plan. The fairness can be indicated
in terms of both criteria, such as equality of total time cut
or equality of total power cut. Moreover, fairness is also
considered between each appliance in the same household.
The EMS provider aims to keep fairness for each appliance
in the same household, except when users declare their preference for the appliances. Thus, we also design two appliance
preference factors in terms of both criteria and integrate them
respectively according to the following user inconvenience
weight factor:
1) Time Cut Method

X 
jnew = j 1 + j 1t
li,j ,
(3)
i
new
=
i,j

new
i,j
P new ,
i i,j

(4)

new
i,j
= i,j 1 + i,j 1t li,j ,

(5)

2) Power Cut Method



X
jnew = j 1 + j 1t
pi,j li,j ,

(6)

i
new
i,j

new
i,j
P
=
new ,
i i,j

new
i,j
= i,j 1 + i,j 1t pi,j li,j ,

(7)
(8)

where 1t is the time slot, j is the plan parameter, and i,j is

the preference parameter. Therefore, the EMS provider can
ensure fairness for users subscribing to the same plan and
each appliance by updating the user inconvenience weight
factor and appliance preference factor after performing DRM.
C. CONVEX RELAXATION

Considering tht problem (2) is an integer linear programming (ILP) and that such problem is generally an
NP-hard problem. Although many approaches have been
developed to address ILP, they are still impractical because
they do not guarantee solving the problem in polynomial
time, especially for large systems, such as the proposed
system which consists of a large number of appliances.
To address problem (2), we relax the constraint of li,j {0, 1}
to li,j [0, 1]. Moreover, the selection li,j must satisfy the
2 A similar strategy has proven useful in [30] for the application of electric
vehicle charging.
VOLUME 3, 2015

W.-T. Li et al.: DRM for Residential Smart Grid

constraints of the real status. For example, if si,j = 1, then

li,j can be 1 or 0. By contrast, if si,j = 0, then li,j can
only be zero. Consequently, the optimization problem can be
approximated as
X X
min
j
i,j li,j
li,j , i,j

s.t.

D
E+
pi,j li,j PU Ppeak

i,j

0 li,j si,j ,

i, j

(9)

The optimization problem is a linear programming (LP) and

thus a convex optimization problem. To solve the problem,
we employ CVX, a package of MATLAB for specifying and
solving convex programs [31].
The relaxed problem is not generally equal to the original problem because the optimal results of the relaxed
? , can be fractional. Thus, l ? should be mapped
problem, li,j
i,j
into 0 or 1 by using a threshold to obtain the desired results,
which can be performed as
(
1, if lji? ,
lji? =
(10)
0, if lji? < ,

the algorithm proceeds to the next time slot and repeats

from step 1.
V. TESTBED AND EXPERIMENTAL RESULTS

In this section, we describe the testbed focusing on the residential DRM. We then perform a DRM experiment on our
testbed to demonstrate the effectiveness and efficiency of the
proposed plans and methods.
A. TESTBED SETUP

The overall architecture of the proposed system is shown

in FIGURE 1; it mainly consists of of three layers: third party
EMS, cloud server, and households. We have built the testbed
shown in FIGURE 2, which consists of the three modules, to
match the proposed architecture. The details of the implementations are described in the subsequent subsections.

where is a given threshold. In our experience, = 0.5 can

generally yield good results. Hence, we adopt such setting
(i.e., = 0.5) in our experiments.
Algorithm 1 Control Algorithm for DRM
initialize: j (0) = 1, i,j (0) = 1/nj , i, j, t = 1
input : the plan parameter j , and the preference
parameter i,j
e
peak (t), i, j;
1 Read si,j (t), pi,j (t), Pj (t), P
U
2 Compute P (t) by using (1);
3 Solve Problem (9) by using CVX;
?
4 Decide lji i, j by using (10);
5 Weightupdating:
Timecutmethod:
Update user inconvenience weight factors and
appliance preference factors by using (3)(5).
Powercutmethod:
Update user inconvenience weight factors and
appliance preference factors by using (6)(8).
6
7

t = t + 1;
Repeat to 1;

The implementation procedure of the proposed control algorithm is summarized in Algorithm 1. In short,
in steps 1 and 2 of the algorithm, the EMS provider
computes total the demand power and reads the threshold given by the grid. In step 3 and 4, the EMS provider
decides the optimal demand status by solving problem (9).
In step 5, the EMS provider updates user inconvenience
weight factors and appliance preference factors based on
the method it uses, that is (3)(5) or (6)(8). Finally,
VOLUME 3, 2015

FIGURE 2. Demonstration of testbed module.

1) CLOUD SERVER

The cloud server is where we can store data of households and

provide services for both third party management services
and households. In our testbed, we implement a cloud server,
which consists of a home gateway manager and database, and
select XMPP and RESTful HTTP as the network protocol to
be used in the proposed system. The extensible messaging
and presence protocol (XMPP) is an application profile of
the extensible markup language (XML), which enables the
near-real-time exchange of structured yet extensible data
between multiple network entities [32]. In the proposed system, the control commands must also be pushed toward the
end users for different applications, which require an asynchronous communication model [33]. XMPP is capable not
only of sending asynchronous request but also of supporting a
massive number of users at the same time through its pub/sub
protocol [34]. Hence, XMPP is selected to send control commands from the home gateway manager of the cloud to all the
gateways of all the households. In addition, RESTful HTTP
is lightweight, has a simple HTTP request format, and is
2435

W.-T. Li et al.: DRM for Residential Smart Grid

TABLE 2. Description of test-bed equipment.

very easy to implement. Moreover, RESTful HTTP is best

suited for applications that require periodic communication.
Hence, RESTful HTTP is selected for the periodic uploading
of sensor data from the gateway of households to the database
of the cloud.

on the solution results to the home gateway manager of the

cloud server. We summarize the above descriptions and the
equipment of the testbed in Table 2.

FIGURE 4. Demonstration of (a) base load demand and (b) probability of

remaining turned on of the controllable loads in the considered system.

B. EXPERIMENTAL RESULTS
FIGURE 3. Household module of testbed.

2) HOUSEHOLDS

We also build a home gateway (HG) and several smart plugs

for each HG to simulate a household, as shown in FIGURE 3.
HG is responsible for two tasks: 1) collecting data of appliances and uploading to the database of the cloud server and
2) receiving control command from the cloud server and
sending the control command to the respective devices.
HG, which is an IP-based system and possesses a two-way
communication channel, the HTTP and XMPP protocols, can
serve as a translator and communicate with non-IP based
devices, such as the smart plugs in the system. In our testbed,
the Raspberry pi computer (Model-B Rev 1) serves the role
of HG and is connected to the z-wave smart plug through
a RaZberry module, as shown in FIGURE 2. Smart plugs
can be equipped with different appliances to send the data of
energy usage to HG and perform the control command from
HG (e.g., switch on/off).
3) THIRD PARTY MANAGEMENT SERVICES

Finally, a PC is used as third-party management service so

that EMS is implemented based on MATLAB 2014b. Such
EMS can access data from the database of the cloud server
by RESTful HTTP protocol and perform DRM, which solves
the optimization problem (9) through the CVX package of
MATLAB. Then, EMS sends the control command based
2436

In our experiment, we use the testbed to simulate a small

residential community consisting of ten households, with
each household equipped with one HG and two smart plugs.3
Lighting appliances are considered to be flexible loads and
equipped with smart plugs to be controllable with two kinds
of states: on and off. Each household then contains two lighting appliances, with the light bulb range between 500 watts
to 200 watts; we always place the light bulb with the higher
rating as appliance 1. We assume the total essential load of the
households as the base load. The base load model, as shown in
FIGURE 4(a), is assumed to be a random curve based on the
reports of National Electricity Market of Singapore (NEMS)
and is used to generate the daily base load curve. We generate
the daily usage patterns of the flexible loads based on their
probability of being turned on, as shown in FIGURE 4(b).4
In FIGURE 5, we show how the proposed system
can use the system to control the light appliances in
each household to reduce the peak demand in a smart
grid. According to FIGURE 5, a duration of 24 hours
is divided into 8640 time slots, that is, the smart grid
detects the total demand every 10 seconds. The green
and blue zones of the figure denote the load profile with
3 The system is designed such that the number of households and number
of appliances can be scaled easily without the need of changing the algorithm.
4 We generate the base load and the daily usage patterns of the flexible
loads by MATLAB. We then send the commands in terms of generated usage
patterns to smart plugs to simulate the users behavior.
VOLUME 3, 2015

W.-T. Li et al.: DRM for Residential Smart Grid

FIGURE 5. Demonstration of peak load shaving through controlling loads

of the homes using the proposed architecture.
FIGURE 7. Demonstration of result without appliances preference
(power cut method).

FIGURE 6. Demonstration of result without appliances preference

(time cut method).
FIGURE 8. Demonstration of result with appliances preference (time cut
method with less DRM over appliance 2).

and without demand management, respectively; the gray

zone indicates the base load, and the red line indicates
the maximum allowable peak demand threshold (33 kW).
As demonstrated in FIGURE 5, when the total demand is
under the peak limit, the smart grid does not need to do
any controlling; hence, no difference is observed between the
green zone and blue line. However, once the total demand
exceeds the peak limit, the system controls the lights and
turns some of them off to reduce the demand load promptly as
indicated by the blue line from 9:00 to 18:00 hours. Thus, the
result in FIGURE 5. clearly shows that our testbed can effectively perform EMS, DRM, and other sevice applications.
To demonstrate the effectiveness and efficiency of the proposed plans and method, we assume that households 1 to 4
adopt GAP and others agree with the GSP. Therefore, we
set j = 3, j = 1, . . . , 4 and j = 1, j = 6, . . . , 10
for both methods. Moreover, we also assume that users do
not declare their preferences of appliances. We then set the
default value for all the appliances, that is, i,j = 1, i, j.
We simulate the usage patterns of one month and show
the result in the FIGURE. 6-9, which contains four subfigures. The first row from left to right describes the
total off-time of the household and the total off-time of
the individual appliances. The second row from left to right
describes the total cut power of the household and the total
cut power of the individual appliances.
VOLUME 3, 2015

FIGURE 9. Demonstration of result with appliances preference (power

cut method with less DRM over appliance 2).

First, EMS performs DRM based on the time cut method,

and FIGURE 6 shows that the total off-time of the GAP
households is lower than those that subscribe to GSP. Moreover, the ratio of the off-time, which is around 1 : 3, conforms
with the proposed plans. The off-time is almost equal for
households that adpot the same plan. Similarly, the off-time of
each appliance is almost equal in the household. By contrast,
EMS performs DRM based on the power cut method, and
2437

W.-T. Li et al.: DRM for Residential Smart Grid

TABLE 3. MSE of total power load w/wo DRM during 9-18 oclock.

FIGURE 7 shows that the total cut power of the GAP

households is lower than those that subscribe to GSP. Moreover, the ratio of the cut power, which is around 1 : 3,
conforms with the proposed plans. The amount of cut
power is almost equal for households that adpot the same
plan. Similarly, the amount of cut power of each appliance is almost equal in the household. Given that the
load of lighting appliance 2 is less than that of lighting appliance 1, the total off-time of lighting appliance
2 is higher than that of lighting appliance 1 to meet the
criteria.
Next, we consider the users preferences for appliances
and assume that all users prefer less DRM for appliance 2.
We then set 1,j = 1 and 2,j = 2, j with both
methods. FIGURE 8 shows the results of DRM in terms
of the time cut method. The total off-time of the households is similar to that in FIGURE 6, but the total offtime of appliance 1 is around twice that of appliance 2.
Similarly, FIGURE 9 show the results of DRM in terms
of the power cut method. The total of the cut power of
the households is similar to FIGURE 7, but the total cut
power of appliance 1 is around twice that of appliance 2.
FIGURE 8-9 demonstrate that the proposed methods can
allow the user preferences.
To demonstrate the performance of the proposed methods,
we introduce the mean square error (MSE) as a performance
metric, which is defined as
n
o
MSE1 = E kPU Ppeak k22 ,
(11)
n
o
MSE2 = E khPU Ppeak i+ k22 .
(12)
In Table 3, MSE1, MSE1 is the result of MSE from
9:00 to 18:00 hours of one month, and MSE2 is the result
of MSE, which measures only those situations whenever
the total power load exceeds the threshold Ppeak from
9:00 to 18:00 hours of one month. For the case of the total
load without DRM, MSE1 and MSE2 are equal. This is due
to the total load without DRM, which always exceeds the
threshold Ppeak from 9:00 to 18:00 hours. Moreover, the MSE
of the total load without DRM is far more than the results
for all the cases of the total load with DRM. This shows the
effectiveness of the proposed DRM scheme. A different also
exists between the simulation and testbed results because of
the delay in the testbed, where the smart meter is sampled
at intervals of 10 seconds. Thus, a delay occurs when the
EMS receives the data, performs optimization, and sends the
control. However, no such delay occurs in the simulation.
In addition, the performance of the power cut method is
slightly better than that of the time cut method.
2438

VI. CONCLUSIONS

In this paper, we propose two types of customer engagement plan, namely, GSP and GAP, which describe two user
inconvenience indices of participation with DRM and the
incentive. Furthermore, we develop an appropriate DRM algorithm with two methods, in terms of user inconvenience
indices, to facilitate EMS performing DRM on peak load
reduction. The objective of the proposed algorithm is to
determine the demand states of all the appliances of all
households subject to the total demand load below a given
threshold.
We build a testbed to verify the proposed scheme and
compare its performance versus its simulation. The results are
useful to evaluate the gap between actual implementation and
theoretical study, which is an important step to produce a better design that is more suitable for real-world implementation.
In addition, the proposed algorithm can also be extended in
a decentralized fashion, such that EMS assigns each user an
individual threshold according to his subscription plan, and
users can individually determine the demand states of their
appliances.
REFERENCES
[1] M. Erol-Kantarci and H. T. Mouftah, Wireless sensor networks for costefficient residential energy management in the smart grid, IEEE Trans.
Smart Grid, vol. 2, no. 2, pp. 314325, Jun. 2011.
[2] M. Rastegar and M. Fotuhi-Firuzabad, Outage management in residential
demand response programs, IEEE Trans. Smart Grid, vol. 6, no. 3,
pp. 14531462, May 2015.
[3] N. Ul Hassan, Y. I. Khalid, C. Yuen, and W. Tushar, Customer engagement plans for peak load reduction in residential smart grids, IEEE Trans.
Smart Grid, vol. 6, no. 6, pp. 30293041, Nov. 2015.
[4] L. Chuan and A. Ukil, Modeling and validation of electrical load profiling
in residential buildings in Singapore, IEEE Trans. Power Syst., vol. 30,
no. 5, pp. 28002809, Sep. 2015.
[5] N. Forouzandehmehr, M. Esmalifalak, H. Mohsenian-Rad, and Z. Han,
Autonomous demand response using stochastic differential games,
IEEE Trans. Smart Grid, vol. 6, no. 1, pp. 291300, Jan. 2015.
[6] A. Albert and R. Rajagopal, Thermal profiling of residential energy use,
IEEE Trans. Power Syst., vol. 30, no. 2, pp. 602611, Mar. 2015.
[7] Y. Liu, C. Yuen, R. Yu, Y. Zhang, and S. Xie, Queuing-based energy
consumption management for heterogeneous residential demands in smart
grid, IEEE Trans. Smart Grid, Jun. 2015.
[8] X. Wang, C. Yuen, X. Chen, N. Ul Hassan, and Y. Ouyang, Cost-aware
demand scheduling for delay tolerant applications, J. Netw. Comput.
Appl., vol. 53, pp. 173182, Jul. 2015.
[9] M. T. Bina and D. Ahmadi, Stochastic modeling for the next day domestic
demand response applications, IEEE Trans. Power Syst., vol. 30, no. 6,
pp. 28802893, Nov. 2015.
[10] M. Ahmadi, J. M. Rosenberger, W.-J. Lee, and A. Kulvanitchaiyanunt,
Optimizing load control in a collaborative residential microgrid environment, IEEE Trans. Smart Grid, vol. 6, no. 3, pp. 11961207, May 2015.
[11] A. Anvari-Moghaddam, H. Monsef, and A. Rahimi-Kian, Optimal smart
home energy management considering energy saving and a comfortable lifestyle, IEEE Trans. Smart Grid, vol. 6, no. 1, pp. 324332,
Jan. 2015.
VOLUME 3, 2015

W.-T. Li et al.: DRM for Residential Smart Grid

[12] Y. Liu, C. Yuen, S. Huang, N. Ul Hassan, X. Wang, and S. Xie, Peakto-average ratio constrained demand-side management with consumers
preference in residential smart grid, IEEE J. Sel. Topics Signal Process.,
vol. 8, no. 6, pp. 10841097, Dec. 2014.
[13] C. Chen, J. Wang, and S. Kishore, A distributed direct load control
approach for large-scale residential demand response, IEEE Trans. Power
Syst., vol. 29, no. 5, pp. 22192228, Sep. 2014.
[14] A. Safdarian, M. Fotuhi-Firuzabad, and M. Lehtonen, A distributed
algorithm for managing residential demand response in smart grids,
IEEE Trans. Ind. Informat., vol. 10, no. 4, pp. 23852393,
Nov. 2014.
[15] C. Bartusch and K. Alvehag, Further exploring the potential of residential demand response programs in electricity distribution, Appl. Energy,
vol. 125, pp. 3959, Jul. 2014.
[16] C. Vivekananthan, Y. Mishra, G. Ledwich, and F. Li, Demand response
for residential appliances via customer reward scheme, IEEE Trans.
Smart Grid, vol. 5, no. 2, pp. 809820, Mar. 2014.
[17] X. Fang, S. Misra, G. Xue, and D. Yang, Smart gridThe new and
improved power grid: A survey, IEEE Commun. Surveys Tuts., vol. 14,
no. 4, pp. 944980, Oct. 2012.
[18] P. Siano, Demand response and smart gridsA survey, Renew. Sustain.
Energy Rev., vol. 30, pp. 461478, Feb. 2014.
[19] J. Aghaei and M.-I. Alizadeh, Demand response in smart electricity
grids equipped with renewable energy sources: A review, Renew. Sustain.
Energy Rev., vol. 18, pp. 6472, Feb. 2013.
[20] J. S. Vardakas, N. Zorba, and C. V. Verikoukis, A survey on demand
response programs in smart grids: Pricing methods and optimization
algorithms, IEEE Commun. Surveys Tuts., vol. 17, no. 1, pp. 152178,
Mar. 2015.
[21] Y. Zhang, S. He, and J. Chen, Data gathering optimization by dynamic
sensing and routing in rechargeable sensor networks, IEEE/ACM Trans.
Netw., Jun. 2015.
[22] I. Al-Anbagi, M. Erol-Kantarci, and H. T. Mouftah, Delay critical smart
grid applications and adaptive QoS provisioning, IEEE Access, vol. 3,
no. 1, pp. 13671378, Aug. 2015.
[23] M. Erol-Kantarci and H. T. Mouftah, Energy-efficient information and
communication infrastructures in the smart grid: A survey on interactions and open issues, IEEE Commun. Surveys Tuts., vol. 17, no. 1,
pp. 179197, Mar. 2015.
[24] R. C. Qiu et al., Cognitive radio network for the smart grid: Experimental
system architecture, control algorithms, security, and microgrid testbed,
IEEE Trans. Smart Grid, vol. 2, no. 4, pp. 724740, Dec. 2011.
[25] V. Salehi, A. Mohamed, A. Mazloomzadeh, and O. A. Mohammed,
Laboratory-based smart power system, part I: Design and system
development, IEEE Trans. Smart Grid, vol. 3, no. 3, pp. 13941404,
Sep. 2012.
[26] W.-Z. Song, D. De, S. Tan, S. K. Das, and L. Tong, A wireless smart
grid testbed in lab, IEEE Wireless Commun., vol. 19, no. 3, pp. 5864,
Jun. 2012.
[27] C. Wang et al., A highly integrated and reconfigurable microgrid
testbed with hybrid distributed energy sources, IEEE Trans. Smart Grid,
Oct. 2014.
[28] F. Adinolfi, G. M. Burt, P. Crolla, F. DAgostino, M. Saviozzi, and
F. Silvestro, Distributed energy resources management in a low-voltage
test facility, IEEE Trans. Ind. Electron., vol. 65, no. 4, pp. 25932603,
Apr. 2015.
[29] W. Tushar, C. Yuen, K. Li, K. L. Wood, W. Zhang, and X. Liu, Design of
cloud-connected IoT system for smart buildings on energy management,
EAI Trans. Ind. Netw. Intell. Syst., Oct. 2015.
[30] C.-K. Wen, J.-C. Chen, J.-H. Teng, and P. Ting, Decentralized plug-in
electric vehicle charging selection algorithm in power systems, IEEE
Trans. Smart Grid, vol. 3, no. 4, pp. 17791789, Dec. 2012.
[31] M. Grant and S. Boyd. (Mar. 2014). CVX: MATLAB Software for
Disciplined Convex Programming, Version 2.1. [Online]. Available:
http://cvxr.com/cvx
[32] P. Saint-Andre, Streaming XML with Jabber/XMPP, IEEE Internet
Comput., vol. 9, no. 5, pp. 8289, Sep./Oct. 2005.
[33] H. Flores and S. Srirama, Mobile cloud messaging supported by
XMPP primitives, in Proc. 4th ACM Workshop Mobile Cloud Comput.
Services (MCS), Taipei, Taiwan, Jun. 2013, pp. 1724.
[34] M. Kirsche and R. Klauck, Unify to bridge gaps: Bringing XMPP into
the Internet of Things, in Proc. IEEE Int. Conf. Pervasive Comput.
Commun. Workshops (PERCOM Workshops), Lugano, Switzerland,
Mar. 2012, pp. 455458.
VOLUME 3, 2015

WEN-TAI LI (S12) received the B.S. and

M.S. degrees in optoelectronics and communication engineering from National Kaohsiung Normal
University, Kaohsiung, Taiwan, in 2009 and 2011,
respectively. He is currently pursuing the Ph.D.
degree with the Institute of Communications Engineering, National Sun Yat-sen University and is
a research assistant with the Singapore University
of Technology and Design, SUTD, Singapore. His
research interests lie in power system monitoring,
the smart grid, and wireless communications.

CHAU YUEN (SM12) received the B.Eng. and

Ph.D. degrees from Nanyang Technological University, Singapore, in 2000 and 2004, respectively. He was a Post-Doctoral Fellow with Lucent
Technologies Bell Laboratories, Murray Hill,
in 2005. He was a Visiting Assistant Professor
with Hong Kong Polytechnic University in 2008.
From 2006 to 2010, he was with the Institute
for Infocomm Research, Singapore, as a Senior
Research Engineer. He has been with the
Singapore University of Technology and Design as an Assistant Professor
since 2010.
He received the IEEE Asia-Pacific Outstanding Young Researcher Award
in 2012. He serves as an Associate Editor of the IEEE TRANSACTIONS ON
VEHICULAR TECHNOLOGY.

NAVEED UL HASSAN (M08SM15) received

the B.E. degree in avionics engineering from the
College of Aeronautical Engineering, Risalpur,
Pakistan, in 2002, and the M.S. and Ph.D. degrees
in electrical engineering, with a specialization
in digital and wireless communications, from
the Ecole Superieure dElectricite, Gif-sur-Yvette,
France, in 2006 and 2010, respectively. In 2011,
he joined as an Assistant Professor with the
Department of Electrical Engineering, Lahore
University of Management Sciences, Lahore, Pakistan. Since 2012, he has
been a Visiting Assistant Professor with the Singapore University of Technology and Design, Singapore. He has several years of research experience.
He has authored/co-authored numerous research papers in refereed international journals and conference proceedings. His major research interests
include cross-layer design and radio resource optimization in wireless networks, demand response management in smart grids, indoor localization, and
heterogeneous networks.

WAYES TUSHAR (S06M13) received the

B.Sc. degree in electrical and electronic engineering from the Bangladesh University of Engineering and Technology, Bangladesh, in 2007, and
the Ph.D. degree in engineering from Australian
National University, Australia, in 2013. He was a
Visiting Researcher with National ICT Australia,
ACT, Australia. He was also a Visiting Student
Research Collaborator with the School of Engineering and Applied Science, Princeton University, NJ, USA, in summer 2011. He is currently a Research Scientist with
the Singapore University of Technology and Design (SUTD)MIT International Design Center, SUTD, Singapore. His research interests include signal
processing for distributed networks, game theory, and energy management
for smart grids. He was a recipient of two best paper awards, both as the first
author, in the Australian Communications Theory Workshop 2012 and the
IEEE International Conference on Communications 2013.
2439

W.-T. Li et al.: DRM for Residential Smart Grid

CHAO-KAI WEN (M04) received the

Ph.D. degree from the Institute of Communications Engineering, National Tsing Hua University,
Hsinchu, Taiwan, in 2004. He was with the Industrial Technology Research Institute, Hsinchu, and
MediaTek Inc., Hsinchu, from 2004 to 2009. He is
currently an Associate Professor with the Institute of Communications Engineering, National
Sun Yat-sen University, Kaohsiung, Taiwan.
His research interests center around the optimization in wireless multimedia networks.
KRISTIN L. WOOD received the B.Sc. degree
in engineering science from Colorado State University, Fort Collins, CO, USA, in 1985, and the
M.Sc. and Ph.D. degrees in mechanical engineering from the California Institute of Technology,
Pasadena, CA, USA, in 1986 and 1989, respectively. He joined as a Faculty Member with The
University of Texas at Austin in 1989, after completing his Ph.D. work, and established a computational and experimental laboratory for research
on engineering design and manufacturing, in addition to a teaching laboratory for prototyping, reverse engineering measurements, and testing. From
1997 to 1998, he was a Distinguished Visiting Professor with the United
States Air Force Academy (USAFA), where he worked with the USAFA
Faculty to create design curricula and research within the Engineering
Mechanics/Mechanical Engineering Department. In 2011, he was a Professor
with the Mechanical Engineering, Design and Manufacturing Division, The
University of Texas at Austin. He was a National Science Foundation Young
Investigator, the Cullen Trust for Higher Education Endowed Professor in
Engineering, a University Distinguished Teaching Professor, and the Director of the Manufacturing and Design Laboratory and MORPH Laboratory.

2440

KUN HU received the B.S. degree in computer science and technology from Jilin University, in 2012,
and the M.S. degree in software engineering from
Peking University, in 2015. He was a Visiting
Student Researcher with the Singapore University
of Technology and Design in 2014. His research
interests lie in Internet of Things, pervasive computing, and smart home.

XIANG LIU received the B.Eng. degree in electrical engineering from Hunan University, PRC
in 1984, the Ms.Eng. degree in pattern recognition and intelligent control from the Institute of
Automation, the Academy of Sciences of China,
in 1987, and the Ph.D. degree in automation from
Bordeaux University 1, France, in 1993.
Starting 1994, Dr. Liu joined the Global Software Group (GSG), Motorola. In 2001, He had
been elected as the Member of Motorola Science
Advisory Board Associates (SABA) for his technical contribution. Starting
March 2003, Dr. Liu joined the School of Software and Microelectronics, Peking University, as a Professor and the Chair of the Department
of Embedded System Engineering. His current research interests include
high performance embedded computing, ubiquitous computing, and software
engineering.

VOLUME 3, 2015