(IJACSA) International Journal of Advanced Computer Science and Applications,

Vol. 15, No. 10, 2024

Advanced Techniques for Optimizing Demand-Side

Management in Microgrids Through Load-Based
Strategies
Ramia Ouederni1, Bechir Bouaziz2, Faouzi Bacha3
Computer Laboratory for Electrical Systems, LR11ES26, INSAT, University of Carthage, Tunisia 1, 2, 3
Electronics & Telecommunications Department, ISIMG, University of Gabes, Tunisia 2
Electrical Department, ENSIT, University of Tunis, Tunisia3

Abstract—Microgrids are crucial for ensuring reliable electrical infrastructure built around locally available renewable
electricity in remote areas, but integrating renewable sources like energy technologies at considerably reduced costs. In addition,
photovoltaic (PV) systems presents challenges due to supply they can accelerate access to electricity in areas where it is
intermittency and demand fluctuations. Demand-side impossible to extend the grid in the short or long term.
management (DSM) addresses these issues by adjusting
consumption patterns. This article explores a DSM strategy To ensure that the electrical system becomes more reliable
combining load shifting (shifting demand to periods of high PV and robust, the peak demand is considered as opposed to the
generation), peak clipping (limiting maximum load), and valley average demand. As a result, environmental resources are being
filling (redistributing load during low-demand periods). wasted and production systems are being under-used. Rapid-
Implemented in MATLAB and tested on a PV-battery microgrid, response power generators (e.g. gas and oil-fired units),
the strategy significantly reduces peak demand, improves the employed to satisfy the demand peak, not only cost a lot of
peak-to-average demand ratio (PAR), and enhances system money, they also generate high levels of carbon emissions. As a
stability and flexibility, particularly with the inclusion of response, various packages were presented to help users shape
deferrable loads. their energy use profiles. These programs aim to make efficient
use of the generation available so that new transmission and
Keywords—Demand side management; microgrid; load shifting; generation infrastructure is installed as little as possible. Such
peak clipping; valley fill
programs, referred to as demand-side management (DSM)
I. INTRODUCTION programs, are designed to program consumption or reduce it [6].
In recent decennials, an evolution in conventional power grid A DSM program provides support for power grid
structure has taken place due to uncertain distributed renewable functionalities in various areas, such as electricity market
generation (URG) and advanced communication and control control, infrastructure maintenance, and management of
technologies [1], [2]. The increased utilization of deep decentralized energy resources [7]. In electricity markets, it
penetration URGs and the improvement of demand-side informs the load controller about the latest load schedule and
resources (DSR) development have been seen as two important possible load reduction capabilities for each time step of the next
technological advances for supporting the sustainable expansion day. Using this procedure, it schedules the load according to the
of today's energy infrastructure, enabled by the objective of total objectives of interest associated with the power distribution
cost reduction [3]. A microgrid (MG) that can be incorporated systems [8], [9]. The load shapes indicate industrial or
with various URGs, advanced energy storage systems (ESS), residential consumers' daily or seasonal electricity demands
and DSRs represents a sustainable way to deliver renewable, between peak hours (PHs) and off-peak hours (OPHs). These
clean electricity as a vital part of the power transformation of shapes can be modified by six techniques [9], [10]: peak
conventional electricity production systems [4]. clipping, valley filling, load shifting, strategic conservation,
strategic load growth, and flexible load shape.
Microgrids (MGs) are stand-alone systems that support the
integration of distributed power resources (DERs) and Both peak clipping and valley filling are techniques for
renewable energies for both reliability and profitability [4]. They controlling loads directly. While peak shaving is concerned with
are important components of intelligent energy transmission and peak load reduction, valley filling takes into consideration load
distribution networks that regulate the electrical system's shaping on off-peak loads. The best and widely employed
resilience under unfavorable conditions. method of managing load in today's electricity networks is load
shifting. It involves shifting the load from the PH to the OPH. In
The various solutions for supplying rural areas with the case of conservation strategy [10], demand reduction
electricity are extensions of the national grid, autonomous methods are used at the customer's end to obtain optimal
systems, or microgrids [5]. Microgrids, or small electrical charging patterns. If the load requirement is greater, the daily
networks, have become a major solution because they can response is optimized using distributed energy resource load
function independently. They are essential to developing an augmentation techniques [11], [12].

1211 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

The principal aims of the DSM are to remodel a load profile  Substantial power consumption savings can be realized
and regulate the highest and lowest day-to-day power through corresponding generation and load demand
requirements to make the most of power sources and prevent the requirements without deep-discharging of battery
need to construct new generation power plants. This may result storage.
in the consumption of electricity. Many DSM technologies are
designed to maximize efficiency to avoid or reduce the need to  The global load profile is enhanced by exploiting
build more power plants. Further reasons to implement DSM available renewable energy resources to a large extent,
techniques include social relations and environmental thus minimizing total reliance on the storage system.
considerations, by modifying customers' power consumption The remaining sections in this paper are structured in the
patterns at peak and off-peak times to conserve power [13]. following way. In Section II, we describe the configuration of
In this paper, three DSM methods were applied for a the microgrid, in Section III we present the DSM method and its
microgrid of model photovoltaic solar panels with real measured features, and in Section IV we present the DSM problem.
demand as well as radiance data to show the benefits of energy Section V presents the detailed results and discussion, while
efficiency which is currently lacking on the island of Djerba, Section VI concludes the paper.
Tunisia. A model system was simulated in the MATLAB II. CONFIGURATION OF A MICROGRID SYSTEM
software using the DSM load-carrying capability. The principal
research contributions of this work are summarized in detail A schematic representation of the proposed microgrid
below. system is shown in Fig. 1. This system comprises three different
groups of units. The first includes energy production units with
 The DSM methods of load shifting, load capping, and photovoltaic panels. The second is comprised of energy
valley filling were applied on a microgrid model of the consumption units, principally residential electrical loads. The
island of Djerba, Tunisia. third is dedicated to energy storage, as represented by a typical
battery bank.

Fig. 1. Schematic representation of the proposed microgrid system.

III. DEMAND SIDE MANAGEMENT free of charge, making them easy to implement in development
regions [14]. According to the [15] definition of demand-side
Demand side management (DSM), also known as energy management, DSM comprises three main concepts: demand
demand management, consists of implementing strategies and response, energy efficiency, and energy conservation.
procedures to improve the efficiency of an energy system on the
supply and demand side. It enables us to manage, control, The demand response (DR) approach involves modifying
influence, and generally minimize power consumption, giving customer electricity consumption, shifting load from periods of
us a better perspective on the potential of renewable energy peak demand to off-peak times, and using active demand pricing
systems, or making effective use of those systems that have been strategies: by modifying electricity prices over time (when it's
installed. Such strategies can provide substantial financial off-peak, it's less expensive) or through inducement payments,
advantages to energy end-users and reduce carbon emissions. these offer customers an incentive to reduce their demand for
The measures aimed at reducing energy consumption do not energy at a particular period of the day. DR is usually linked to
necessarily have to be costly, and can often even be provided advanced energy metering technologies that provide
information to end-users on their consumption of energy.

1212 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

DSM can be implemented in several ways, and Fig. 2 shows price variations [21]. They are seen as an effective solution to
several categorizations of DSM with associated load shapes. some of the problems of deregulated power systems.
Peak clipping aims at reducing demand during peak hours.
Utilities achieve this control by incentivizing customers not to Demand response enables industrial, commercial, and
consume power during peak hours, directly controlling loads, or residential customers to optimize their electricity consumption
setting up higher prices. The method is helpful in cases with no to obtain reasonable prices and improve grid reliability. In other
possibility of setting up or installing new power plants [16]. words, it can modify electrical energy consumption to reduce
Valley filling focuses on raising usage during periods of very peak loads and shift consumption to off-peak hours [22].
low electricity profile to keep demand and supply balanced, Demand response methods generally fall into two categories:
avoiding the startup and ramp-up costs of generators [17]. Load electricity price-based programs and incentive-based programs.
growth is more common when using electric vehicles, where Fig. 3 illustrates this classification.
customers are encouraged to increase usage up to a certain In this article, among the various demand management
threshold for grid stability [18]. Load shifting gives consumers programs, the load shifting program has been used to reduce
options to shift their usage pattern to off-peak hours based on peak load and increase load during periods of low demand (i.e.
cheap tariffs. peak load reduction and load trough filling). There are also
It is the combination of load clipping and valley filling [19], restrictions on time-of-use shifting for each of the loads.
[20]. Flexible load shaping is when consumers are flexible
enough to shift their loads to different low-usage slots. Usually,
customers willing to participate in this are identified and
incentivized for their participation [16]. Energy efficiency is
when the overall load profile is lowered throughout the day by
using more energy-efficient devices or through cyclic operation
[21].
A demand curve is formed by various DSM technologies
[21], as illustrated in Fig. 2, additionally, but without limitation:
 Load shifting: The main objective of load shifting is the
transfer of load from peak to off-peak hours.
 Conservation: The best-known approach to reduce global
consumption of electricity, and not only at peak periods.
 Peak clipping: The main aim of this strategy is to reduce
demand at peak times (e.g. 7 p.m.). Energy consumption
is reduced by the control of equipment that can be
switched off, like air conditioners or heaters, by the
customer or the electricity supplier.
 Valley fill: Under valley fill, the demand for load Fig. 2. Classification strategies for demand-side management.
increases in off-peak hours and reduces during peak
hours.
 Load growth: Unlike the conservation approach, with
load expansion, the overall amount of electricity sold
(off-peak and on-peak) increases.
 Load Shape Flexibility: With this technique, the utilities
have the power to discontinue loading as required, and
without notifying the customer. Changes in the quality or
quantity of service are generally described as flexible
load shapes.
Practically all the DSM programs are designed and
implemented to optimize the use of existing power plants.
Another crucial objective is to avoid, defer, or delay the
construction of much-needed new power plants (both renewable
and conventional).
A. Demand Response
These programs are recent developments in demand-side
management, involving consumers' participation in improving Fig. 3. Different types of demand response programs.
their energy consumption patterns in response to instantaneous

1213 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

To illustrate the process of implementing the Demand Side A. System Modeling Equations
Management (DSM) method, the following pseudo-algorithm 1) Power generated by the photovoltaic system: The output
describes the key steps in the approach used:
generated by the photovoltaic panels, PPV (t), is given in the
Algorithm 1: Pseudo-code of the DSM proposed algorithm equation below:
Step 1
Enter data, including renewable energy production (solar PPV (t) = ηPV × APV × Irr (t)
irradiance), load demands, system parameters (PV area, yield), and Where:
threshold values for demand-side management (DSM).
Step 2 ηPV: representing the efficiency of the photovoltaic panels.
Calculate hourly energy production using input data and system
parameters. APV: represents the overall area of the panels in m2.
Step 3
Initialize load curve for DSM operations (load shifting and peak
Irr(t): represents the solar radiation in W/m2 at the time t.
shaving). 2) Peak clipping strategy: The peak reduction is achieved
Step 4
Identify periods with significant energy production.
by setting a maximal limit Pseuil above which the load output is
Determine the time of maximum energy production. restricted:
Step 5
For each period of the load curve: Pload_clipped (t) = min(Pload (t), Pseuil )
If the load exceeds the DSM threshold:
Where:
Calculate excess load above a threshold.
Reduce the load to a threshold level. Pload(t): is the initial load at the current time.
Redistribute excess load to adjacent periods with sufficient energy
production (Load Shifting). Pseuil: is the limit value selected.
Step 6
For each period of the modified load curve: 3) Load shifting strategy: The load shift consists of shifting
If the load still exceeds the threshold: the surplus charge from the peak times to the times when the
Limit load to threshold level using Peak Clipping. photovoltaic generation is highest. If at a given time t, the
Step 7 Pload(t) load is above the limit, the Pexces(t) surplus is determined
Apply the Valley Fill method:
Identify periods of low demand (valleys) in the load curve. as described below:
Redistribute non-critical load from high-demand periods (after Load
Pexces (t) = Pload (t) − Pseuil 
Shifting and Peak Clipping) to these low-demand periods to smooth
the load curve. This excess is allocated to the hours of most significant PV
Step 8 generation (defined as more than 50% of the maximum PV
Repeat steps 5, 6, and 7 until the load curve is optimized. generation):
Step 9
Return the optimized load curve and analyze the results. Pexces (t)
Pload_shifted (h) = Pload_shifted (h) + 
End NPV

IV. PROBLEMATIC DEMAND MANAGEMENT SYSTEM Where:

For a microgrid system that includes the use of photovoltaic h: denotes a time selected for redistribution.
(PV) generation, optimization of power consumption is a crucial NPV: is the total number of hours selected for redistribution.
factor in reducing costs and maximizing the utilization of
renewable resources. Two strategies commonly used to meet 4) Valley-filling strategy: A valley fill approach involves
this objective are demand-side management (DSM): the redistribution of under-utilized capacity at times of low
demand (off-peak hours) to cover the valleys in the load curve.
 Peak Clipping: The purpose of this strategy is to reduce
the maximum power requirement by clipping and If the load Pload(t) is less than a predetermined level at a given
reducing loads above a specific level. time t, the load deficit Pdeficit(t) is calculated in the following
equation:
 Load Shifting: With this approach, loads are shifted
during peak periods to periods when photovoltaic Pdeficit (t) = Pseuil − Pload (t)
generation is most available, thus minimizing To smooth demand, this deficitis then assigned to the periods
dependence on conventional sources. of lowest photovoltaic production (defined as less than 50% of
 Valley Fill: In this method, the loads in periods when peak photovoltaic output):
demand is low are increased to fill the valley of demand, Pdeficit (t)
thus leveling out demand and optimizing the use of Pload_filled (h) = Pload_filled (h) + 
Nlow_PV
available resources, thereby minimizing variations and
dependence on traditional sources of energy.

1214 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

Where: capacity of 889.128 kW. This remote microgrid is designed to

supply power to the island's communities, contributing to the
h: represents a period selected for the redistribution. region's energy independence and sustainability. The analysis
Nlow_PV: represents the number of total hours selected for the was conducted over a week during which irradiance
redistribution when the photovoltaic output is low. measurements were taken, to assess the load profile and
irradiance patterns to evaluate the solar power generation of the
5) Mathematical formulation of optimization: The general system. Fig. 4 illustrates the power consumption over this
objective of the problem is to minimize the peak loads and period, Fig. 5 presents the irradiance data, and Fig. 6 represents
maximize the utilization of renewable energy. The problem can the photovoltaic system’s power output. Comparing these
be expressed in the following terms: results with similar studies in microgrid systems reveals
consistent trends in power production and consumption patterns,
min (maxPload_final (t)) particularly in remote island contexts. Studies conducted in
other island communities show similar variability in solar
With the constraints below for peak clipping: irradiance and load profiles. However, our findings suggest that
the Djerba microgrid demonstrates higher efficiency in
Pload_final (t) ≤ Pseuil 
converting solar energy, likely due to optimized PV system
Excess load is redistributed in a balanced manner during load design and favorable climatic conditions. This has important
shifting, with consideration of PV generation. This approach consequences for isolated microgrid design and operations,
ensures that load adjustments align with solar energy underlining the necessity for an enhanced energy control
production, optimizing efficiency and reducing reliance on non- strategy. In comparison to earlier research, this study
renewable sources underscores the potential for enhancing load-shifting techniques
and optimizing storage systems, which could further reduce
V. RESULTS AND DISCUSSION reliance on backup generators and improve overall energy
A. Basic Case efficiency in isolated microgrids.
This study utilized a dataset from a microgrid located on the
island of Djerba in Tunisia, with a maximum solar power

Fig. 4. Weekly load power analysis.

Fig. 5. Weekly analysis of solar irradiation levels.

1215 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

Fig. 6. Weekly performance analysis of photovoltaic power system.

B. Effect of Peak Clipping Fig. 7 illustrates the impact that peak clipping has on the
A controller or switching equipment carries out the peak hourly load. The blue curve shows the initial load, with
clipping process. The consumption is continuously supervised significant daily fluctuations. The second curve represents the
and when it rises to levels that give reason for concern, the load after peak shaving, with load values restricted to a 10-kW
controller can switch off specific loads to minimize the demand. threshold. The peak clipping process limits the peak
The loads that can be disconnected are ranked in order of consumption to a specified threshold, thereby preventing system
priority, and classified as transferable and non-transferable. surges and stabilizing the energy demand. It effectively holds
the load below a monitored level, thus minimizing the potential
for power system instability and overloading.

Fig. 7. Impact of peak clipping DSM on load profile.

C. Effect of Load Shifting curve represents the load after implementing load shifting,
Fig. 8 presents an example of demand-side management where consumption peaks have been effectively smoothed out.
(DSM) implementation through load shifting. An analysis of the This method redistributes surplus load to hours of higher
results indicates that load shifting has been strategically applied photovoltaic production, resulting in a more uniform demand
over 24 hours, with most operations scheduled during peak solar distribution. Comparative analysis with similar studies shows
irradiation to minimize the need for energy storage. By aligning that load shifting is a widely recognized strategy for optimizing
load utilization with times of highest production, the system energy consumption in microgrids, demonstrating its
maintains a sufficient margin to supply critical loads during peak effectiveness in reducing peak loads and alleviating pressure on
demand periods, thereby enhancing system stability and the power supply system. Our findings corroborate previous
preventing overloads. The impact of load shifting on hourly load research, highlighting that load shifting not only enhances
is illustrated in Fig. 8. The initial load, shown in blue, reveals system reliability but also supports sustainability efforts by
significant peaks during high-demand periods, while the second maximizing the utilization of renewable energy resources.

1216 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

Fig. 8. Impact of Load Shifting DSM on 24-Hour Load Profile.

D. Effect of Valley Fill Fig. 9 shows the effect of the Valley Fill method on the load
The Valley Fill approach is implemented in the form of an per hour. The initial blue curve represents the load, characterized
intentional increase in load during times of reduced demand, by steep peaks and troughs during certain periods of the day. The
usually monitored by an appropriate controller or energy other curve shows the load after Valley Fill has been applied,
management system. As consumption falls to a critical level, a with the troughs filled to smooth demand. This stabilizes power
controller may enable supplementary loads, prioritized based on demand by preventing excess fluctuations, thus reducing the risk
their ability to transfer loads without affecting the global system. of under-utilizing the available energy resources and improving
The loads that can be shifted are increased to compensate for any global system efficiency.
dips in the power curve.

Fig. 9. Load Profile Analysis Before and After Valley Fill DSM.

Shifting reduces power demand and inconvenience in peak shifting enables off-peak hours to be rescheduled with a smaller
hours by switching surplus demand away from the system. Load energy consumption reduction, as illustrated in Fig. 10.

1217 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

Fig. 10. The impact of DSM strategies on load profile: load shifting, peak clipping, and valley fill.

VI. CONCLUSION [7] J. Yang, W. Xu, K. Ma, J. Chen, W. Guo, “Integrated demand-side
management for multi-energy system based on non-cooperative game and
This study examines energy management in a solar-powered multi-energy pricing,” Sustainable Energy, Grids and Networks, vol.34,
microgrid using the Demand Side Management (DSM) method. 2352–4677,2023.
The DSM approach was applied to a 24-hour solar microgrid on [8] R. Ouederni, B. Bouaziz and F. Bacha, “Design and evaluation of an
the island of Djerba, Tunisia, utilizing data from NASA. The island's hybrid renewable energy system in Tunisia,” 2022 5th
International Conference on Advanced Systems and Emergent
results show that after implementing DSM strategies, the load Technologies (IC_ASET), Hammamet, Tunisia, 2022, pp. 418-423,2022.
profile aligned more closely with solar generation, as much of
[9] [Link]-Porras, A.López, J.C. Salazar-Elena, “Deregulation in the
the energy demand was shifted to periods of higher solar output. Energy Sector and Its Economic Effects on the Power Sector: A Literature
This approach provides a framework for achieving more Review,” Sustainability, vol. 13, 3429,2021.
resilient isolated microgrids by optimizing the use of renewable [10] [Link], [Link], [Link], [Link], “A Survey of Efficient
energy. Additionally, reducing energy consumption in the Demand-Side Management Techniques for the Residential Appliance
microgrid enhances system stability and offers cost savings. By Scheduling Problem in Smart Homes,” Energies, vol.15, 2863,2022.
improving the balance between load and generation, the method [11] B. Williams, D. Bishop, [Link], J.G. Chase, “Demand Side
also reduces the need for extensive storage capacity, especially Management in Industrial, Commercial, and Residential Sectors: A
Review of Constraints and Considerations,” Energies, vol.16, 5155,2023.
during periods of low solar power availability.
[12] [Link], [Link], and F. Bacha, “A Case Study of Hybrid
REFERENCES Renewable Energy System Optimization for an Island Community based
on Particle Swarm Optimization,” Engineering, Technology & Applied
[1] [Link], S. Peyghami, A. Sangwongwanich, F. Blaabjerg, “Reliability Science Research. vol.14, 3, 14367–14373,2024.
aspects in microgrid design and planning: Status and power electronics- [13] S. N. Bragagnolo, J. C. Vaschetti, and F. Magnago, “A technical and
induced challenges, Renewable and Sustainable Energy Reviews,” economic criteria comparison on demand side management with multi-
vol.159, pp.1364–0321,2022. level optimization model,” IEEE Latin America Transactions 19(9):
[2] Bouaziz, B., Bacha, F. Direct power decoupling control using optimal 1494–1501,2021.
switching table combined with virtual flux estimator for PWM rectifier. [14] R. Ouederni, B. Bouaziz, F. Bacha, “ Modeling and cost optimization of
Electr Eng 106, 1605–1617 (2024). [Link] an islanded virtual power plant: A case study of Tunisia,” Turk J
02179-6 ElectrPower Energy Syst, 2(2), 168–179, 2022.
[3] [Link], and X. Xia,“Techno-economic and environmental optimization [15] [Link], [Link], [Link], [Link], [Link], Demand Side
of a household photovoltaic-battery hybrid power system within demand Management for the Energy-Efficient Future of Pakistan Using a Low
side management,” Renew. Energy 108, 132–143,2017. Emission Analysis Platform, Engineering Proceedings, vol.
[4] [Link], Z. Eskandar Filabi, [Link], “Techno-Economic Analysis 46(1):28,2023.
and Optimization of an Off-Grid Hybrid Photovoltaic–Diesel–Battery [16] S. Panda, S. Mohanty, [Link], [Link], [Link], [Link],
System: Effect of Solar Tracker,” Sustainability, vol.14, 7296,2022. [Link], “Residential Demand Side Management Model, Optimization
[5] [Link], [Link], [Link], “Techno-economic analysis of hybrid solar- and Future Perspective: A Review,” Energy, vol.8, 3727–3766,2022.
diesel-grid connected power generation system, ” Journal of Electrical [17] [Link], [Link], [Link], [Link], [Link], [Link],
Systems and Information Technology,2017. “Exploiting Deep Learning for Wind Power Forecasting Based on Big
[6] S. Nanibabu, S. Baskaran, P. Marimuthu, “Realization of Demand Side Data Analytics,” Applied Sciences, vol. 9(20):4417,2019.
Management in Smart Grid,” Optimization Techniques for Hybrid Power
Systems, 378–406, 2024.

1218 | P a g e
[Link]
 (IJACSA) International Journal of Advanced Computer Science and Applications,
Vol. 15, No. 10, 2024

[18] S. Maharjan, Y. Zhang, S. Gjessing, D.H.K. Tsang, “User-Centric Algorithms,” International Journal of Advanced Computer Science and
Demand Response Management in the Smart Grid with Multiple Applications, vol.15(7), 2024.
Providers,” IEEE Trans. Emerg. Top. Comput, vol.5, 494–505,2017. [21] [Link], D. Dietrich, “Demand Side Management: Demand Response,
[19] [Link], [Link] Jande, [Link], “Demand-Side Intelligent Energy Systems, and Smart Loads,” IEEE Trans. Ind. Inform,
Management of Solar Microgrid Operation: Effect of Time-of-Use 7, 381–388,2011.
Pricing and Incentives,” J. Renew. Energy 2020, 6956214,2020. [22] [Link], [Link], “Integration scenarios of Demand Response into
[20] R. Ouederni, B. Bouaziz, F. Bacha, “Optimization of a Hybrid Renewable electricity markets: Load shifting, financial savings and policy
Energy System Based on Meta-Heuristic Optimization implications,” Energy Policy vol. 96, 231–240,2016.

1219 | P a g e
[Link]