An Incentive-based Optimal Energy Consumption Scheduling Algorithm for Residential Users

Abstract

Smart Grid is the most promising concept which is more reliable, flexible, controllable and environment friendly. Home energy management (HEM) system is an important part of the smart grid that provides a number of benefits to the end users such as savings in the electricity bill, reduction in peak demand and meeting the demand side requirements. Demand Response (DR) and Time-of-Use (ToU) pricing refer to programs which offer incentives to the end users who curtail their energy use during times of peak demand. This paper proposes an energy efficient optimization model based on Binary Particle Swarm Optimization (BPSO) for residential electricity consumers. The proposed model optimally schedules the electricity consumption of different household appliances in a dynamic pricing environment to benefice the user by minimizing electricity cost. Simulation results illustrate that the proposed method efficiently shifts the appliances operation time from high peak to low peak hours and leads to significant electricity bill saving.

Full text

Available online at www.sciencedirect.com
Procedia Computer Science 00 (2015) 000–000
www.elsevier.com/locate/procedia
6th International Conference on Ambient Systems, Networks and Technologies, ANT 2015 and
the 5th International Conference on Sustainable Energy Information Technology, SEIT 2015
An Incentive-based Optimal Energy Consumption
Scheduling Algorithm for Residential Users
Ihsan Ullaha, Nadeem Javaidb,∗
, Zahoor A. Khanc,
Umar Qasimd, Zafar A. Khane, Sahibzada A. Mehmooda
aUniversity of Engineering &Technology Peshawar, Pakistan
bCOMSATS Institute of Information Technology, Islamabad, Pakistan
cCIS, Higher Colleges of Technology, Fujairah Campus, UAE
dUniversity of Alberta, Alberta, Canada
eMirpur University of Science and Technology, Mirpur Azad Kashmir, Pakistan
Abstract
Smart Grid is the most promising concept which is more reliable, flexible, controllable and environment friendly. Home energy
management (HEM) system is an important part of the smart grid that provides a number of benefits to the end users such as
savings in the electricity bill, reduction in peak demand and meeting the demand side requirements. Demand Response (DR) and
Time-of-Use (ToU) pricing refer to programs which offer incentives to the end users who curtail their energy use during times of
peak demand. This paper proposes an energy efficient optimization model based on Binary Particle Swarm Optimization (BPSO)
for residential electricity consumers. The proposed model optimally schedules the electricity consumption of different household
appliances in a dynamic pricing environment to benefice the user by minimizing electricity cost. Simulation results illustrate that
the proposed method efficiently shifts the appliances operation time from high peak to low peak hours and leads to significant
electricity bill saving.
c
2015 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Conference Program Chairs.
Keywords: Appliance scheduling; Binary Particle Swarm optimization; Energy management System; Electricity Pricing; Smart Grid.
1. Introduction
Future smart grid has been considered as an intelligent electricity generation, transmission and delivery system
equipped with an advanced information and control technologies. It aims at improving the efficiency and reliability
of the grid, and relieving economic and environmental issues caused by the traditional fossil-fueled generation [1].
Global energy demand increases steadily each year while the growth of electrical energy generation and transmission
setups increases at a much slower rate. Therefore currently, increasing the total generation capacity is a function of
∗Corresponding author. Website: www.njavaid.com; Tel.: +92 300 5792728.
E-mail address: nadeemjavaid@comsats.edu.pk; nadeemjavaidqau@gmail.com
1877-0509 c
2015 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the Conference Program Chairs.

2Author name /Procedia Computer Science 00 (2015) 000–000
peak demand. Efficient supply-demand management and great exploitation of renewable energy are two important
features of a smart grid. Due to its limited capability in information-exchange, the traditional grid suffers from
inefficient operations at both the supply-side and the demand-side [2].
According to the U.S consumers electricity data report, household appliances consume about 42% of the residential
energy [3]. Energy optimization is one of the active topic and one of the major challenge that needs a proper attention
this time. Different optimization approaches, models and schemes for better energy management and consumption
is proposed and deployed. New devices based on new technologies are being deployed e.g., controllable household
appliances, advanced smart meters, stand-alone electrical energy generation and storage systems, i.e., plug-in hybrid
electric vehicle batteries and a communication infrastructure of high potentials. The shift in the energy consumption
pattern of the household appliances form high peak hours to low peak hours is said to be Demand Response (DR).
The shift causes an increase in the energy demand at that specific time horizon [4], [5].
Price based demand response programs consider flattening demand fluctuations as their ultimate objective. Both
the customer and the utility will benefit from DR [6]. A DR strategy coordinates the requirements and needs between
the energy provider and the customer [7]. The end users that take part in the DR program get benefits from it, but
besides it DR program is also beneficial for the utility grid. The DR program provides benefits to the utility grid by
protecting it from different risk factors like blackouts, thus increasing the smart grid reliability. It also reduces the
peak demand at certain time horizon, thus reduce the need of the expensive generation plants [8].
2. Related Work
Recently, residential energy management has become an active topic with respect to research and also has a need
of an implementation on the real test bed. In EM system, appliance scheduling is one of the main and important
parameter that needs proper attention and therefore, several appliance scheduling strategies have been proposed by
different researchers. Smart grid is a network of technologies that provides electricity from power plants to the end
user and connects all supply, grid and demand elements via an effective communication system.
In [9], EM system is deployed in a home to schedule the electricity consumption in such a way that PAR and
electricity Cost is reduced to the maximum extent. An optimization approach based on RTP combined with the IBR
pricing scheme is used for the power consumption of all the Automatically Operated Appliances (AOA) in the home.
The objective of the DSM strategy is to increase the use of renewable energy resources, increase the economic
benefit and reduce the power imported from the main distribution grid or minimize the peak load demand [10].
According to the proposed architecture, the objective load curve is taken as an input by the DSM system and demands
for the control action in order to meet the desired load consumption.
A Genetic Algorithm (GA) based optimization approach combined with a two point estimate method is used to
meet the Heating Ventilation and Air conditioning (HVAC) load with a hybrid renewable energy generation and energy
storage system. Hybrid generation systems are inherently unpredictable because of the intermittent nature of the wind,
solar irradiance [11].
Residential users are not much aware about the importance of DR program and most of them have not tools for
taking part in DR. Since, residential customers have not resources, a few residential users take part in a DR program.
Commercial and industrial customers have tools and therefore widely participate in the DR programs [12].
Distributed resources are used to describe mainly three new concepts, i.e., DR, distributed generation and elec-
tricity storage. These distributed resources are connected in low and medium voltage level inside of the grid. This
connection of distributed resources inside of the grid represents a radical change for the operation [13]. An oppor-
tunistic scheduling scheme is proposed based on the optimal stopping rule for smart appliance automation control
[13], [3].
The rest of the paper is organized as follows. The motivation of the paper is discussed in section III. In section
IV, the system model is discussed and the problem is formulated as an optimization problem. Simulation results are
presented in section V to show the performance and comparison with traditional users and smart users. Finally we
concluded in section VI.

Author name /Procedia Computer Science 00 (2015) 000–000 3
3. Motivation
Smart grid gives opportunity to the end users to communicate bi-directional with the utility in real-time, so con-
sumers can tailor their energy consumption based on individual preferences like price concern, etc. Based on the
different usage pattern of the electrical devices, the Smart Grid offers dynamic pricing scheme in order to avoid dif-
ferent risk factors like a blackout or load shedding, thus allows the user to curtail the energy consumption during high
peak hours. Mostly, the users are not aware of the DR program and also have no tools to take part in a DR program,
thus may not consume the energy optimally and pay maximum cost. Based on the dynamic pricing, there is therefore a
need to develop smart systems that will autonomously execute all these tasks without the prompting of the customers.
In this paper, we proposed a demand side energy management model for a household that is connected to the grid
and also generate some energy from RES. The model is efficient, smart, robust and is capable of coping with the
price uncertainty and maximizes the utility of the customers by consuming energy optimally and benefices the user
by paying minimum electricity cost. In this model, the SS receives the dynamic price signal from the grid and adjusts
the hourly load level of the user in response to hourly electricity prices. The SS firstly, schedules the smart appliances
by shifting the maximum allowable load from high peak hours to low peak hours. Secondly, the SS checks the hourly
energy cost and switches the load from the smart grid to the RES storage where the load costs maximum.
4. Proposed Approach to optimize the energy consumption
In this section, an optimal approach for scheduling the power usage of smart appliances in home is proposed based
on the ToU pricing scheme. In this work, we present a new and an innovative model to anticipate the electricity
usage pattern for residential electrical appliances. The HEM system is equipped with an intelligent Smart Schedular
(SS) and all the household appliances bi-directionally communicate with the SS. The home has rooftop Renewable
Energy Source (RES) generation and storage system. In this model, the SS finds the operation pattern for the smart
appliances in the search space and decides the optimal time for the smart appliances in order to benefice the resident
by minimizing the electricity cost. The SS optimally utilizes the utility grid energy as well as RES stored energy.
During low peak hours, the SS utilizes the grid energy and shifts the load from grid to the RES system when the grid
energy costs maximum to the user. The simulation results reveal that the presented appliance scheduling scheme offer
benefits to the household and generally enjoy lower bills as compared to the users having no HEM architecture in
their homes. Moreover, the energy consumers are categorized based on the scheduling pattern of the appliances, i.e.,
i) Traditional users- this class of user is taken into account without HEM. This type of users are non-price sensitive,
ii) Smart users- this class of user is taken into account with HEM architecture, and iii) Smart Prosumers- this type
of users not only consume the grid energy but also produce some energy from RES. This class of users has HEM
architecture and RES both in their homes.
An optimal approach for scheduling the power usage of smart appliances in home is proposed based on the ToU
pricing scheme. An algorithm based on BPSO technique is used to anticipate the optimal time for making the appli-
ances to operate. All the home appliances communicate with the central control unit, i.e., SS at each time horizon h.
The SS based on the BPSO calculates and generates the optimal time pattern for all the household appliances. The SS
checks the 24 hour time horizon and take decisions based on that optimal generated pattern to operate the appliances
in order to complete the task.
4.1. Problem Formulation for Appliance Scheduling
This section presents how BPSO can solve the appliance scheduling problem. First, we define the optimization
problem following the modified implementation of the BPSO algorithm, which in combination with RES cognition,
provide promising performance.
4.2. The Optimization Problem
It is of great importance to distribute loads properly in the Hhour horizon, such that one can get the maximum
profit out of the smart in-home system. Thus, we define the optimization problem as follows:

4Author name /Procedia Computer Science 00 (2015) 000–000
Fig. 1: Appliance power rating
Given a set of appliances, i.e., A={a1,a2, ..., aN}, where each appliance consumes different energy and their
consumption rating is shown in fig. 1. Such appliances are connected to the SS of the HEM. The cost of electricity in
the smart grid is based on the time, i.e., different prices are set by the utility for different time horizon over a day. We
assumed four levels of ToU, namely high peak cost, shoulder peak, low peak cost, offpeak, and is tabulated in table.1.
It is common that during high peak and low peak hours the electricity price is relatively high and low, respectively.
Noticeably, each user uhas one goal, to optimally consume the energy by paying the minimum electricity cost.
The overall objective function is to minimize its electricity bill payment and is hence formulated as follows:
min
24
∑
h=1
Ch(1)
subject to :
N
∑
a=1
24
∑
h=1
Eh,a≤Egrid (2)
1≤h≤24 (3)
where Cis the cost of energy and Eh,ais the amount of energy consumed by the appliance aduring time horizon h.
Particle swarm optimization is a heuristic population based search technique that locates the solution to an opti-
mization problem. Optimization involves binary-valued so Binary Particle Swarm Optimization (BPSO) technique is
used to find the best fitness value for the objective function. Particles represent candidate solutions in a solution space
and the optimal solution is found through moving the particles in the D-dimensional solution space. The particles
initial positions and velocities are randomly initialized. Nparticles combine together to form a swarm. Afterwards,
they move around the solution space to find the optimal solution. The global best at the end of the simulation is taken
as the solution to the problem. The fitness of all particles are evaluated and the global and personal best positions are
updated if needed. Each particle then fly in the search space and each particle dynamically update their position and
velocity by tracking two extremes, i.e., Plbest and Pgbest in each iteration.
5. Results and Discussions
To evaluate the performance of the proposed appliance schemes, we simulated the daily energy use of a set of
household appliances. The attributes, i.e., number of appliances and the power rating of the appliances were set as
shown in fig. 1. Simulations are performed for three main cases i.e., i) traditional user, ii) smart user and iii) smart

Author name /Procedia Computer Science 00 (2015) 000–000 5
prosumer. This study assumes that a household PhotoVoltaic (PV) generation must be able to meet at least 30% of its
load demand. The ToU pricing policy is used for billing of the energy users. These prices are typically established
Table 1: ToU pricing scheme
High Peak hours Shoulder peak hours Low peak hours Offpeak hours
1am-4am,7pm-9pm 9am-2pm 5am-8am,10pm-12am 3pm-6pm
well in advance by the utility grid. The differential or ToU pricing provides financial incentives to the customers who
take part in DR program for shifting their load from high peak to offpeak periods. In differential pricing, the cost of
electricity is charged at different rates during different time horizons of the day and is tabulated in table.1. In order
0 5 10 15 20 25
2
3
4
5
6
7
8
9
10
11
12
Time(hours)
Energy consumption (KWh)
without HEM
(a) Unscheduled load energy consumption.
0 5 10 15 20 25
0
20
40
60
80
100
120
140
Time (hours)
Cost (Rs)
without HEM
(b) Unscheduled load energy cost.
Fig. 2: Traditional User profile.
to demonstrate the effectiveness of our different designed appliance scheduling schemes, the simulation results and
comparison of all the three cases and their performance are analyzed and discussed in this section.
The differential time pricing policy is used for billing of the traditional energy users. They, therefore wholly rely
on the utility grid to meet the power demands of their electrical devices. The energy obtained from the grid and is
consumed by different appliances in different time horizon is shown in fig. 2a and the energy cost for the unscheduled
load is shown in fig. 2b. In the second case, the home has smart appliances and HEM system is improved with an
intelligent Smart Scheduler (SS), so called smart homes. These are the customers who have no RES and therefore
wholly rely on the utility grid to meet the power demands of their electrical appliances. The daily energy consumption
profile of the smart home is shown in fig. 3a.
The SS efficiently responds against the utility tariffand avoids the appliances to operate during high peak hours
and thus benefices the user to pay minimum electricity bill. For scheduled load, a plot of daily cost is demonstrated
in fig. 3b. It is evident from fig. 3b that the SS shifts the appliances pattern from high peak hours to shoulder, low
and offpeak hours which results in minimizing the end user electricity bill. The energy consumed by the appliances
in smart homes in comparison with unscheduled load is demonstrated in fig. 4a. It is evident from fig.4a that the SS
shifts the appliances efficiently from high peak hours to low peak hours. The daily energy cost of both the smart user
and the traditional user is shown in fig. 4b.
In the third scenario, the users have a smart appliances scheduler as well as RES and storage system. Such user
daily generates 30% of the energy of its total daily load. The performance of the algorithm to optimally consume the

6Author name /Procedia Computer Science 00 (2015) 000–000
0 5 10 15 20 25
2
4
6
8
10
12
14
16
Time(hours)
Energy consumption (KWh)
with HEM
(a) Energy consumption.
0 5 10 15 20 25
10
20
30
40
50
60
70
80
90
100
110
Time(hours)
Electricity Cost (Rs)
with HEM
(b) Scheduled load energy cost.
Fig. 3: Smart User profile.
0 5 10 15 20 25
2
4
6
8
10
12
14
16
Time(hours)
Energy consumption (KWh)
with HEM
without HEM
(a) Unscheduled and scheduled load energy consumption.
0 5 10 15 20 25
0
20
40
60
80
100
120
140
Time(hours)
Cost (Rs)
with HEM
without HEM
(b) Unscheduled and scheduled load energy cost.
Fig. 4: Comparison of Traditional and Smart user.
grid energy as well as the RES is shown in fig. 5a. The SS utilizes the RES stored energy and shifts the load from the
grid to RES stored energy and thus minimizes the electricity cost by a very significant amount. The energy cost of the
SS with PV generation is shown in fig. 5b. From the fig. 5a, it is clear that the SS optimally schedules the appliances
where EP is minimum and shifts the maximum possible load to RES storage system during high peak cost. In this
way the resident exploits optimally the RES stored energy during high peak cost and eliminates the high peaks in the
electricity cost. The cost profile of the HEM enabled home is quite minimum as compared to the traditional users. The
energy cost is Rs.1404.5 for the home without HEM, the cost is Rs.1132.5 for the home with HEM which accounts
for about 19.36% of reduction in the smart user bill.
Finally, a relative comparison of the three cases is taken into account. The energy consumed by the three cases and
the cost against these consumptions is shown in fig. 5a and fig. 5b, respectively. The smart prosumer gets benefits of
the ToU program and thus optimally exploits the grid energy and residential energy as well. For the home, having
HEM with/without RES, total daily energy costs are Rs.804 and Rs.1132.5, respectively. The reduction in energy
consumption cost is approximately 29% in that case. From the simulation results, it is clear that the proposed algorithm
incites the prosumer by 43% every day with respect to traditional user.

Author name /Procedia Computer Science 00 (2015) 000–000 7
0 5 10 15 20 25
0
5
10
15
Time (hours)
Energy consumption (KWh)
HEM with RES
HEM without RES
Without HEM
(a) Energy consumption profile.
0 5 10 15 20
0
20
40
60
80
100
120
Time (hours)
Cost (Rs)
HEM with RES
HEM without RES
Without HEM
(b) Energy cost profile.
Fig. 5: Daily energy consumption and cost profile of the three users.
6. Conclusion
In this paper, a new appliance scheduling model based on BPSO is proposed. The proposed model based on a ToU
pricing scheme efficiently schedules the household appliances and benefices the end user by minimizing the daily
electricity cost. The results obtained from several case studies, including HEM and RES revealed that the model
efficiently schedule the household appliances and cost saving is achieved in the user electricity bill. Simulation results
show that the proposed model and appliance scheduling algorithm reduce the bill of prosumer by more than 29% with
respect to smart users. Moreover, when the house is supplied from the grid only, the HEM architecture proposed in
this paper still reduce the bill by 19.36%.
References
1. T. F. Garrity, “Getting smart”, IEEE Power Energy Mag., May 2009 vol. 8, no. 2, pp. 38â˘
A¸S45.
2. P. P. Varaiya, F. F. Wu, and J. W. Bialek, “Smart operation of smart grid: Risk-limiting dispatch,” Proc. IEEE, Jan. 2011, vol. 99, no. 1,
pp.40â ˘
A¸S56.
3. Yi, Peizhong, Xihua Dong, Abiodun Iwayemi, Chi Zhou, and Shufang Li. “Real-time opportunistic scheduling for residential demand re-
sponse.” Smart Grid, IEEE Transactions on 4, no. 1 (2013): 227-234.
4. Hubert, Tanguy, and Santiago Grijalva. “Modeling for Residential Electricity Optimization in Dynamic Pricing Environments.” Smart Grid,
IEEE Transactions on 3, no. 4 (2012): 2224-2231.
5. Rahimi, Farrokh, and Ali Ipakchi. "Demand response as a market resource under the smart grid paradigm." Smart Grid, IEEE Transactions on
1, no. 1 (2010): 82-88.
6. Ozturk, Yusuf, Datchanamoorthy Senthilkumar, Sunil Kumar, and Gordon Lee. “An intelligent home energy management system to improve
demand response.” Smart Grid, IEEE Transactions on 4, no. 2 (2013): 694-701.
7. Medina, Jose, Nelson Muller, and Ilya Roytelman. "Demand response and distribution grid operations: Opportunities and challenges." Smart
Grid, IEEE Transactions on 1, no. 2 (2010): 193-198.
8. H. Aalami, G. R. Yousefi, and M. P. Moghadam, “Demand response model considering EDRP and TOU programs,” in Proc. IEEE/PES
Transm. Distrib. Conf. Expo. 2008 , pp. 1â ˘
A¸S6.
9. Zhao, Z., Lee, W. C., Shin, Y., & Song, K. B. (2013). “An Optimal Power Scheduling Method for Demand Response in Home Energy
Management System.” Smart Grid, IEEE Transactions on, 4(3), 1391-1400.
10. Logenthiran, Thillainathan, Dipti Srinivasan, and Tan Zong Shun. “Demand side management in smart grid using heuristic optimization.”
Smart Grid, IEEE Transactions on 3, no. 3 (2012): 1244-1252.
11. Arabali, A., M. Ghofrani, M. Etezadi-Amoli, M. S. Fadali, and Yahia Baghzouz. “Genetic-algorithm-based optimization approach for energy
management.” Power Delivery, IEEE Transactions on 28, no. 1 (2013): 162-170.
12. Pipattanasomporn, M.; Kuzlu, M.; Rahman, S.; , “An Algorithm for Intelligent Home Energy Management and Demand Response Analysis,”
Smart Grid, IEEE Transactions on , Dec. 2012,vol.3, no.4, pp.2166- 2173.
13. Mishra, Aditya, David Irwin, Prashant Shenoy, Jim Kurose, and Ting Zhu. “GreenCharge: Managing RenewableEnergy in Smart Buildings.”
Selected Areas in Communications, IEEE Journal on 31, no. 7 (2013): 1281-1293.