Content uploaded by Nadeem Javaid
Author content
All content in this area was uploaded by Nadeem Javaid on Mar 25, 2019
Content may be subject to copyright.
Pro Utility Pro Consumer Comfort
Demand Side Management in Smart Grid
Waleed Ahmad1, Nadeem Javaid1(B
), Basit Karim2, Syed Qasim Jan3,
Muhammad Ali4, Raza Abid Abbasi1, and Sajjad Khan1
1COMSATS University, Islamabad 44000, Pakistan
nadeemjavaidqau@gmail.com
2COMSATS University Islamabad, Abbotabad, Pakistan
3University of Engineering and Technology, Peshawar, Pakistan
4King Saud University, Riyadh, Saudi Arabia
http://www.njavaid.com
Abstract. Now a days, energy is the essential resource and due to
increase in power demand, traditional resources are not enough to ful-
fill the requirement of todays need. The researchers are working on new
approaches to enhance and improve the power load demand. The increas-
ing demand of electricity creates peaks on utility. Therefore an improved
Home Energy Management System (HEMS) is necessary for the automa-
tion of smart home to reduce the cost and peaks on utility. In this paper
work, our objective is pro utility and pro-consumer comfort which means,
thedecreaseinPeaktoAverageRatio(PAR)inordertoreducethe
stress on the utility while increasing user comfort. In this Research, we
have proposed a new technique called Random Cell Elimination Scheme
(RCES) with Demand Side Management (DSM) for a home appliance
scheduling. To make the system more effective, we have utilized two pric-
ing systems: Time of Use (ToU) and Real Time Pricing (RTP) in our
experiment. The simulation results are compared with two heuristic opti-
mization schemes: Bacterial Foraging (BFA) and Firefly Algorithm (FA).
The experimental results shows that the proposed scheme performed 80%
better than BFA and FA in reducing PAR and user discomfort.
1 Introduction
The smart grid is more powerful than the traditional grid system because of bidi-
rectional communication between the source and end users which makes it more
scalable and efficient [1]. The Demand Response (DR) helps in the interaction
between the power source and consumers to provide more efficient and reliable
system [2]. The main role of DR is to reduce the burden on the power source
by motivating the consumers to change their energy usage behavior according
to price tariff provided by a utility. It has been analyzed that more than 40% of
energy is consumed in a residential area around the world [3].
The international energy outlook in 2013 assumes that by 2040, it will rise to
more than 56% [4]. To tackle all these challenges, smart grid emerged with the
c
Springer Nature Switzerland AG 2019
L. Barolli et al. (Eds.): WAINA 2019, AISC 927, pp. 385–397, 2019.
https://doi.org/10.1007/978-3-030-15035-8_36
386 W. Ahmad et al.
concept of more reliability, productivity, and cost-effectiveness with the most
important feature like dynamic control. Demand Side Management (DSM) in
smart grid plays an important role by utilizing optimization techniques in an
adaptable way to sustain the grids balance [5]. The operations of DSM with
Home Energy Management System (HEMS) helps the consumer to control the
load demand dynamically. Fundamentally, the purpose of DSM is to educate the
electricity users to change their usage pattern in order to reduce the cost and bur-
den on the power grid [6]. The DSM strategies include control and management
of load demand and appliance scheduling. However, despite the above-mentioned
strategies, load shifting is used vastly in past research work to manage the load
by DR [7]. This helps the consumers to transfer the load from ON peak to Off-
peak hours to provide incentive to the consumers while lowering the stress on
the main grid to reduce the Peak to Average Ratio (PAR) [8].
The main purpose of the above-discussed approaches is to aware the users to
reduce the PAR and load. Different type of scheduling algorithms is adopted by
DSM to optimize the load demand pattern [9]. As we know, the residential sector
has many household appliances which are running time to time in a whole day,
which attracts the attention of researchers. To overcome this problem, the HEMS
[10] is proposed with the different type of pricing tariffs. HEMS is responsible
to handle and monitor the energy demand in a smart home dynamically [11].
The real-time dynamic scheduling is a difficult task and done by using dynamic
programming [12].
The proposed work of this paper is presented as Follows:
Random Cell Elimination Scheme (RCES): In this paper, our objective is
to present a new heuristic optimization scheme. we analyze the existing heuristic
algorithms by their performance and nature and after evaluating the Bacterial
Forgery Algorithm (BFA) and FireFly Algorithm (FA), we determine that there
need to a more efficient system which can improve multiple limitations such that
reducing the PAR and user discomfort.
2 Related Work
The utilization of smart grid elements by deploying DSM and DR is a demanding
task. In the last few years, many optimization schemes have been proposed and
experimented considering single or multi goals with multiple trade-off objectives.
Some systematic approaches are presented in Table1. In many articles [10,11],
the main focus of research was to reduce the user discomfort, lessen the cost
on the user side and reduction in PAR to benefit the utility directly or by indi-
rectly using some heuristic algorithms, mathematical problems, and statistical
schemes. These types of techniques [12–14], cut down the stress of the power by
transferring the load from ON-Peaks to OFF-Peaks time intervals to reduce the
cost while giving the power to the user to operate a different kind of appliances
according to their preference and comfort. The researchers in [14,15], compose
the problems for multi-objective models, wherein [16], express the problem as a
Pro Utility Pro Consumer 387
single objective optimization. This system schedule the smart devices of an intel-
ligent home. Furthermore, the objective was to stable the load by a mixed integer
linear programming to reduce the cost and PAR. The simulation conducted in
the Czech Republic by taking real-time pricing.
In [14,17], the article focus was to reduce the total cost and reduce the
electricity bill of an individual. Simulations in [18], performed on real-time data
received from Energy and Climate Center UK by taking real time tariffs with
multi-objective mixed integer programming. The high demand determination
schemeisproposedin[15], for residential area using Real Time Pricing (RTP)
for multiple cases. The finite smart appliances are taken with quasi-random
methods and the recursive equation is formed to check the high load demand.
Intelligent HEMS model is presented in [19]. In this model, oU pricing tariff
is used to reduce the total cost without affecting the consumer satisfaction.
In this concern [20], a cost-effective greedy iterative distributed model is used
for multiple consumers to overcome the cost problem in a grid. The simulation
outcomes expose that electricity cost is degraded and PAR in decreased. The
distributed generation model is proposed in [21], which effectively reduced the
cost and waiting time by allowing bidirectional electricity flow. Results show
that for all cases, the 32.73% of energy cost is saved. The user discomfort is
discussed in [22]. In this research paper, the demand response issue is developed
in HEMS with a support of the Markov decision model. The DSM controller is
employed with Genetic Algorithm (GA), Binary Particle Swarm Optimization
(BPSO) and Colony Optimization Algorithm (COA) in [23]. In this scheme, GA
outperforms the rest of the schemes. The cost minimization and waiting time is
directly related to user comfort as discussed in [24].
3 Problem Statement
Traditional grid system has many flaws and limitations. From the last decade,
smart grid attracts the attention of the researcher towards its benefits and advan-
tages. The smart grid contains two-way communication between consumer and
power source. This absolute goal of the smart grid to generate enhancement
within supply, to organize and regulate the whole process of electrical energy
[25]. To meet the requirement of consumers for energy demand by utilizing dif-
ferent techniques identified by DSM. As we know, a single home has too many
electrical appliances which are running time to time with the different type of
power ratings. With the growth in population, power need in the residential side
is raising gradually [26]. By means of employing DSM, a person can help to
eliminate the price as well as handle the unwanted using electrical power by way
of arrangement appliances properly. Through DSM, scheduling of smart devices
is done by knowing the peak hour and estimating the price through the present
hour. These kinds of switching lesson the price but load the utility and therefore
associated with general shortage of energy.
388 W. Ahmad et al.
Table 1. Summarized related work
Technique(s) Achievement(s) Limitation(s)
Linear Programming The proposed scheme
reduce cost and PAR
Only focuses on cost
and PAR, user comfort
is ignored
PSO GA and MKP This model reduces cost
and PAR
UC is ignored
Mixed linear and integer
linear programming
Cost and Peak load is
reduced
User discomfort
increased
Multi-objective mixed
linear programming
Generation cost is
reduced while taking
user comfort
PAR ignored
Double co-operative
game theory
Cost is reduced on both
user and utility side
Smart vehicle charge is
avoided
Game theocratic energy
management
Reduce the load on
utility and user bill
User comfort is not
considered
Quasi random process Load management is
done with recursive
formula
PAR and user comfort
is ignored which burden
the utility
Greedy iterative
algorithm
Cost, PAR reduction
with load shifting
Stress the utility,
increase the energy cost
FA Increase the user
comfort and reduce the
PAR
Cost is high
Reinforcement learning User satisfaction with
reduction in cost
PAR is high
BFA, BPSO and ACO Cost and PAR is
minimized with
improvement in comfort
Renewable energy
resources installation is
not considered
WDO Waiting time is reduced Installation cost is not
observed with energy
cost
4 Proposed Solution
The main focus of our work is to reduce the peak load and total cost which is
calculated in Eq. 1. The RTP and ToU 24-h onward day pricing is used in the
simulation. The central challenging part is load shifting from On-Peak hours to
Off-peak hours in order to tackle to load balance and reduce cost. The load shift-
ing is done based on the fitness criterion which is presented in Eq.2. Now in order
to determine the total load, the Eq. 3is employed, where alpha represents the
status of appliances in 0’s and 1’s. Table 1shows the appliances and their power
rating according to their classifications of interruptible and non-interruptible.
Pro Utility Pro Consumer 389
Cost =
24
hour=1
(Ehour
Rate ∗PApp
Rate) (1)
Ff=min ⎧
⎨
⎩
lieNp
od ≥mean(LUs
od ),Ehour
Rate ≤mean(ERate)
lNp
od >(std(LUs
od )) ∧lieNp
od <mean(LUs
od )
Ehour
Rate >mean(ERate)
(2)
Lod =PApp
Rate ∗App (3)
Our main objective is not only lessened the Cost as defined in Eq. 5but also
decrement in PAR by applying Eq. 6. Along with cost and PAR reduction, we
also have objective to achieve efficient load shifting in DSM as assessed in Eq. 7.
Ojb1=min(Cost) (4)
Ojb2=min(PAR) (5)
Ojb3=min(Load) (6)
To achieve this goal, we have to split a day into 24-h time slots. Furthermore,
these time slots are split into two parts (high price hours and low price hours).
Now the main challenge is to transfer the high load to low load hours according
to their power ratings classification. This approach will effectively decrease the
cost and PAR due to load balance. For calculation of PAR, we have used Eq. 8.
PAR =max(LS
od)
Average(LS
od)(7)
5SystemModel
In our proposed work, simulation is performed on a smart home which consists
of fifteen different appliances that need to be scheduled. Smart appliance Length
of Operation (LoT) and power information is collected from end users. The pro-
posed system model is divided into three sections such as DSM, Supply Side
Management (SSM), and Communication unit. SSM have all the information
related to energy generation and pricing tariff. Energy Management Controller
(EMC) is used by DSM for appliance scheduler. The objective of handling the
load is to bound the consumers from excessive usage of electricity to lower the
burden on utility. The communication unit is responsible for the transfer of
the pricing scheme and dynamic demand-response data between DSM and SSM.
Home Area Network (HAN) is used in a smart home for communication between
the smart appliances and EMC through Wi-Fi and Zig-Bee protocols. Further-
more, there are three classifications of smart devices according to the schedule:
the Interruptible, Un-interruptible, and base load. In our paper, three meta-
heuristic scheme are selected, including FA, BFA and RCES. These schemes are
adopted to manage the smart appliance in HEMS. The central idea of this sys-
tem is to decrease the consumption throughout the peak time to diminish the
cost and PAR. Simulations is performed on actual presumed smart appliances
that are usually split by a few distinct types like: Interruptible, Non-interruptible
and base load (Fig. 1) (Table 2).
390 W. Ahmad et al.
Fig. 1. System model
Table 2. Appliances used in simulations
Appliances Power rating (kwh) LOT (h)
Refrigerator 1.666 24
Vacuum cleaner 0.7 0.7
Water pump 1 8
Washing machine 1.4 3
Cloth dryer 5 4
Dishwasher 1.32 3
Water heater 5 8
Iron 2.4 3
AC 1.5 8
Cooker 0.225 4
Toaster 0.8 1
Printer 0.011 2
Light 0.18 12
Blender 0.3 2
Oven 2.4 4
6 Meta-heuristic Algorithms
Several optimization methods have already been deployed to handle the smart
grid problems. In this regard, several effective heuristic techniques have already
been utilized to produce a strong strategy to tackle energy and smart grid prob-
lems. In that study, we have applied two meta-heuristic systems, BFA and FA.
6.1 BFA
Nature neglects those animals which have bad foraging methods and support
those who have surviving strategies. Initially, BFA algorithm was developed
by [27]. After many productions, poor versions usually are changed or even
Pro Utility Pro Consumer 391
re-designed using healthy versions. The stochastic character of the algorithm
allows the cell to improve statistically and collaboratively approaching the most
effective solution.
6.2 FFA
FFA merge as an obstacle solver [28]. This algorithm helps in giving optimized
solutions to many different optimizations obstacles. Firefly is known as multi-
handler due to its multi-model feature. In our research, we have adopted a new
scheme to tackle the scheduling index.
6.3 RCES
In our research, we have implemented a new Random Cell Elimination Scheme
(RCES). In RCES, random data is generated and the distance of the closest cell
is evaluated. The population is assessed by the number of steps in the defined
position and fitness function is defined to cut back the overall cost. The pro-
duction of the Cell is evaluated and replaced with the fittest position. The new
generation of cell is produced in next step with random population. This new
generation eliminates the old cells based on objective and fitness evaluation. The
main purpose of RCES is o obtain the optimum solution of appliance scheduling
in smart home. However, optimized home scheduling task is a difficult goal to
achieve. Simulation results demonstrate that our system successfully lower the
price and PAR as trade-off between price and delay time, so in order to achieve
one factor, other might be sacrificed.
1. Initialization of the population
2. Fitness Evaluation of each Cell
3. Reproduction
4. Replace previous place with good one.
Algorithm 1. Algorithm for RCES
1: Initialize (PoP,N
p,N
e,N
r,N
c,C
i)
2: Evaluate the initial;
3: for t=1:Nedo
4: for t=1:Nrdo
5: for t=1:Ncdo
6: for t=1:Npdo
7: end For
8: end For
9: end For
10: Calculate Best Solution for task;
11: Pbest=intensity.best with lowest time;
12: Allocate Solution;
13: Evaluate New Solution;
14: end For
15: Find the health(fitness);
16: Select the gbest;
392 W. Ahmad et al.
7 Simulation and Reasoning
In the simulation section, we have selected two heuristic algorithm and evaluation
is done on the foundation of Price, PAR and delay time. Two pricing tariffs ToU
and RTP are utilized in the simulation. The simulation results are evaluated on
two pricing schemes RTP and ToU by using proposed heuristic scheme RCES.
We have used fifteen appliances of smart home that need to be managed and
controlled.
The total price, load per hour and the delay time is denoted in words of cost
and respective time slot. Figures 2and 3represents the total electricity cost in
cents and Kwh of the single smart home by using RTP and ToU price tariff in
term of 24-h time schedule. simulation results shown that each scheme is out-
performing the unscheduled case. The total cost in term of KWh in represented
for 24-h time scheduled. Simulation result reveal that heuristic schemes perform-
ing better than unscheduled. The proposed scheme efficiently reducing the cost
during peak hours as shown in graph.
In Figs. 4and 5, The total load in term of 24 h time interval is shown which
clearly indicate that during peak hour, every heuristic scheme is beating the
unscheduled case. The peak load of unscheduled case is higher while on the
other side heuristic algorithms reduce considerably load in peak times.
Fig. 2. Electricity cost (Cents) TOU
Fig. 3. Electricity cost (Cents) RTP
Pro Utility Pro Consumer 393
Fig. 4. Load (KWh) TOU
Fig. 5. Load (KWh) RTP
The main objective of our approach is to reduce the PAR to lower the burden
on the utility and reduce waiting time. PAR explain the power consumption pat-
tern of the user load and effects the power grids. In Figs. 6and 7, unscheduled and
scheduled PAR is illustrated. The simulation results shows that presented app-
roach RCES reduce considerably amount of PAR as compared to BFA and FA,
which shows that proposed approach is better than other two heuristic schemes.
However, PAR is related to burden on power grids. Simulation results reveal all
the heuristic schemes minimize the PAR. However our proposed heuristic opti-
mization scheme reduce PAR to 80% and beats the other schemes. So we can say
that our new scheme is pro utility and pro user comfort achievement in smart
grid.
Our New proposed scheme also shows the effectiveness in reducing total cost
as compared to unscheduled scenario. All heuristic algorithm shows effective-
ness in reducing the cost. Reducing cost and PAR using heuristic algorithms is
challenging objective because of trade off between cost and delay time. Figures 8
and 9, the results represent the total energy cost for BFA, FA and RCES using
ToU and RTP pricing tariffs.
Figures 10 and 11, present the waiting time which is mainly concern to user
comfort. User comfort is achieved when a consumer faces less waiting time in
operation of electrical appliances. The results shows that our approach is a pro
394 W. Ahmad et al.
Fig. 6. PAR ( To U )
Fig. 7. PAR ( RT P )
Fig. 8. Tot a l c o s t ( To U )
user comfort because of increase in the user comfort by reducing the waiting
time to 50% which is a major contribution of our work.
In order to analyze the model, the confidence level is measured in term of PAR
and waiting time. Simulation results indicates less gap even in random values.
BFA is an old optimization technique which mainly focuses on production and
Pro Utility Pro Consumer 395
Fig. 9. Tot a l c o s t ( RT P )
Fig. 10. Waiting time (ToU)
Fig. 11. Waiting time (RTP)
living of bacteria, while Firefly has known as problem solver because they work
on flashing behavior and distance of the fittest between them. The results reveal
that RCES is efficiently reduce the cost and PAR while scheduling the appliances
towards On-Peak to Off-Peak hour without any stress on the Power Grid.
396 W. Ahmad et al.
8 Conclusion
DSM is used for the control and management of electrical appliances used in
a smart home in order to balance the load demand. In most cases the energy
load is shift to new renewable sources to ease the user but it is not sufficient.
To tackle the above mention problem an optimization technique is adopted. The
new scheme RCES has been proposed which efficiently reduce PAR and user
discomfort. The experiments results clearly revealed that our propose scheme is
better than BFA and FA. The high load is shifted to low peak hours in order
to balance the load and reduce the stress on utility. It is observed that there is
always a tradeoff between cost, PAR and user comfort. The results show that
fitness function improve the performance. The experiment is conducted on the
day ahead ToU and RTP pricing schemes which originally examine the behavior
of optimization schemes. To check the performance of the ToU and RTP pricing
schemes, price forecasting is an essential feature.
References
1. Ipakchi, A., Albuyeh, F.: Grid of the future. IEEE Power Energy Mag. 7(2), 52–62
(2009)
2. Medina, J., Muller, N., Roytelman, I.: Demand response and distribution grid
operations: opportunities and challenges. IEEE Trans. Smart Grid 1(2), 193–198
(2010)
3. Gul, M.S., Patidar, S.: Understanding the energy consumption and occupancy of
a multi-purpose academic building. Energy Build. 87, 155–165 (2015)
4. Khalid, A., et al.: Towards dynamic coordination among home appliances using
multi-objective energy optimization for demand side management in smart build-
ings. IEEE Access 6, 19509–19529 (2018)
5. Gellings, C.W.: The concept of demand-side management for electric utilities. Proc.
IEEE 73(10), 1468–1470 (1985)
6. Broehl, J.H., Huss, W.R., Skelton, J.C., Sullivan, R.W., Limaye, D.R., Sloane,
B.D., Davis, T.D., Blevins, R.P.: Demand-side management, vol. 1. Overview of
key issues. Final report. No. EPRI-EA/EM-3597-Vol. 1. Battelle Columbus Labs.,
OH (USA); Synergic Resources Corp., Bala-Cynwyd, PA (USA) (1984)
7. Davito, B., Tai, H., Uhlaner, R.: The smart grid and the promise of demand-side
management. McKinsey Smart Grid 3, 8–44 (2010)
8. Gelazanskas, L., Gamage, K.A.: Demand side management in smart grid: a review
and proposals for future direction. Sustain. Cities Soc. 11, 22–30 (2014)
9. Logenthiran, T., Srinivasan, D., Shun, T.Z.: Demand side management in smart
grid using heuristic optimization. IEEE Trans. Smart Grid 3(3), 1244–1252 (2012)
10. Shirazi, E., Jadid, S.: Optimal residential appliance scheduling under dynamic
pricing scheme via HEMDAS. Energy Build. 93, 40–49 (2015)
11. Matayoshi, H., Uehara, T., Ludin, G.A., Sharma, A., Datta, M., Senjyu, T.: Unin-
terruptible smart house equipped with a single-phase dq-transformation system.
J. Renew. Sustain. Energy 8(2), 025101 (2016)
12. Mesari, P., Krajcar, S.: Home demand side management integrated with electric
vehicles and renewable energy sources. Energy Build. 108, 1–9 (2015)
Pro Utility Pro Consumer 397
13. Adika, C.O., Wang, L.: Smart charging and appliance scheduling approaches to
demand side management. Int. J. Electr. Power Energy Syst. 57, 232–240 (2014)
14. Gao, B., Zhang, W., Tang, Y., Mingjin, H., Zhu, M., Zhan, H.: Game-theoretic
energy management for residential users with dischargeable plug-in electric vehi-
cles. Energies 7(11), 7499–7518 (2014)
15. Vardakas, J.S., Zorba, N., Verikoukis, C.V.: Power demand control scenarios for
smart grid applications with finite number of appliances. Appl. Energy 162, 83–98
(2016)
16. Zhang, Y., Zeng, P., Li, S., Zang, C., Li, H.: A novel multiobjective optimization
algorithm for home energy management system in smart grid. Math. Probl. Eng.
2015, 19 (2015)
17. Jovanovic, R., Bousselham, A., Bayram, I.S.: Residential demand response schedul-
ing with consideration of consumer preferences. Appl. Sci. 6(1), 16 (2016)
18. Qdr, Q.: Benefits of demand response in electricity markets and recommendations
for achieving them. US Department of Energy, Washington, DC, USA, Technical
report (2006)
19. Abushnaf, J., Rassau, A., Grnisiewicz, W.: Impact on electricity use of introducing
time of use pricing to a multiuser home energy management system. Int. Trans.
Electr. Energy Syst. 26(5), 993–1005 (2016)
20. Chavali, P., Yang, P., Nehorai, A.: A distributed algorithm of appliance schedul-
ing for home energy management system. IEEE Trans. Smart Grid 5(1), 282–290
(2014)
21. Elkazaz, M.H., Hoballah, A., Azmy, A.M.: Artificial intelligent based optimization
of automated home energy management systems. Int. Trans. Electr. Energy Syst.
26(9), 2038–2056 (2016)
22. Wen, Z., O’Neill, D., Maei, H.: Optimal demand response using device-based rein-
forcement learning. IEEE Trans. Smart Grid 6(5), 2312–2324 (2015)
23. Rahim, S., Javaid, N., Ahmad, A., Khan, S.A., Khan, Z.A., Alrajeh, N., Qasim, U.:
Exploiting heuristic algorithms to efficiently utilize energy management controllers
with renewable energy sources. Energy Build. 129, 452–470 (2016)
24. Rasheed, M.B., Awais, M., Javaid, N., Iqbal, Z., Khurshid, A., Chaudhry, F.A.,
Ilahi, F.: An energy efficient residential load management system for multi-class
appliances in smart homes. In: 2015 18th International Conference on Network-
Based Information Systems (NBiS), pp. 53–57. IEEE (2015)
25. California Energy Commission. Energy Efficiency and Conservation: Trends and
Policy Issues. No. 100-03. Report, May 2003
26. Jacob, L., Jeyakrishanan, V., Sengottuvelan, P.: Resource scheduling in cloud using
bacterial foraging optimization algorithm. Int. J. Comput. Appl. 92(1), 14–20
(2014)
27. Passino, K.M.: Biomimicry of bacterial foraging for distributed optimization and
control. IEEE Control Syst. 22(3), 52–67 (2002)
28. Fister, Jr., I., Yang, X.-S., Fister, I., Brest, J., Fister, D.: A brief review of nature-
inspired algorithms for optimization. arXiv preprint arXiv:1307.4186 (2013)
