Smart Grid Optimization: Supply-side and Demand-side Management

This report, submitted as part of an assignment at the University of Wollongong, Australia, explores the optimization of smart grids, focusing on both demand-side and supply-side management. The paper investigates the integration of renewable energy sources, distributed energy storage, and distributed energy generation within the smart grid framework. It highlights the importance of smart meters and demand response mechanisms in achieving optimal power regulation, cost reduction, and user privacy. The report also discusses the challenges and solutions related to voltage regulation, communication protocols, and the integration of distributed generation. Furthermore, it reviews existing literature on real-time pricing, stochastic modeling, and linear programming techniques used to minimize energy costs and improve energy efficiency. The paper proposes the use of Business-to-Grid (B2G) model for optimization, aiming to reduce monetary expenses through energy production and storage rather than solely relying on purchasing from the grid. The report emphasizes the need for advanced control structures, forecasting technologies, and the implementation of Demand-Side Management (DSM) strategies to facilitate a more efficient and sustainable smart grid system.

Optimization for Supply-side and Demand-side Management in Smart Grids
Assignment 1
4/9/2018
University of Wollongong, Australia
PAT
1

Abstract
Demand-Side Management in addition to the integrated distribution of generated energy and its storage, are the ideas being considered to
essentially increase the elements that are used in the implementation of the concept involving smart grid. The renewable resources are not
forgotten in that they are to be balanced as they are massive contributors to today’s power sources. There is a focus on the smart grids in that the
side of demand comprises of traditional users together with users who own some types of energy sources that are distributed as well as some
storage devices for the energy. Using the day-ahead process of optimization, regulation of the power is possible in independent centralized units.
Users who produce some energy are able to help by reducing their expense in monetary energy as they can store as well as produce rather than
only buying form the set grid of power. This paper, therefore, produces a formulated optimized grid with its analysis being based on the existing
optimal strategies. To add on this, the paper suggests the use of smart meters by the power consumers that allow optimal production as well as
storage strategies with a preservation of the user privacy. Such features would lead to a minimized requirement for signalling the central units.
Keywords: Demand-Side Management, smart grid, distributed energy storage, distributed energy storage
2

Introduction
The introduction of smart grids has the important role in the transformation of functionalities in the latest grid s of energy for the aim of
providing user-oriented services as well as an assured high security, economic efficiency and quality of the supplied electricity in the
environment of markets [1]. In addition, the expectations of smart grids are its key enabling feature in the change to low-carbon sectors in energy
thereby ensuring sustainable and efficient consumption of natural resources. Energy production from sources that are renewable that include;
photovoltaic and wind units, however, are naturally intermittent. It, therefore, would mean a lack of correlation in the energy local consumption
and production. Furthermore, the need for dispatchable, flexible is increasing, generation of fast-ramping energy that balances variation in loads
and contingencies that include the transmission losses or the generation of assets [2]. There is a development of these similar problems at the
market level since the local and national balance between demand and supply get complicated when managing greater levels of renewable
energy [1].
Smart structures can be grouped into three major cores, first is the smart protection system that relates to the system’s reliability, security and
failure protection. The targeted optimization areas include the grid-self-healing, information monitoring and data recovery as well as
measurement. Secondly, systems having smart management deal with the utility, cost, energy efficiency and demand profile. Lastly, systems
with smart infrastructure are grouped into sub-classes of smart energy, communication system and information [4].
3

Sources of renewable energy can be integrated with two-way communication existing between the consumer and the grid which is the focused
infrastructure optimization part in the system. Effective demand and supply side need multiple protocols of communication, communications-
enabled with wide area network between the substations and electric utility. Neighbour Area Network, Field Area Network, as well as Advanced
Metering Infrastructure which jointly facilitates communication between consumers and substations. An expensive technology such as the Independent
Power Communication Line is used in optimizing cost efficiency through the combination of simulators in power networks and simulators in communication
networks. Practically, power synchronisation and communication networks are one major problem. Furthermore, advanced infrastructure used in
communication mostly make use of wireless protocols for reasons such as effective work. However, such networks possibly have reduced protection with
their combined communication and power. This could be the leading cause of major damages if a malfunctioning occurs in the system [5].
Penetration in a distributed generation into systems of power distribution support the demand of energy is effective in dealing solving energy
crisis. Also, the systems of energy management are efficient, however, the DG integration is challenging due to problems in the smooth flow of
power. Battery storage and renewable energy systems are integrated with the grid power distribution resulting in unbalancing of power in the
reactive power and active power [6]. The network’s power instability is considerably noticed due to the fluctuations of voltage. This problem can
be corrected with the use of two-way communication method that distributes generation. Systems with two-way communication have intelligent
control strategy that avoids conflict between user devices and provides smooth regulation of voltages [6].
4

The quality of power in the concerning factor in the distribution of power and installation of units produced from renewable energy in the
system’s distribution is affected on the characteristics and functionality in the regulation of voltages. Swags, voltage dips and swell events
featured and could lead to poor quality of power. Traditionally, Step Voltage Regulator, control structures and Feeder Shunt Capacitors are
applied in systems with conventional power distribution. Smart proposals for advanced control structures in the shape of FSC and SVR agents
could reduce the problem of voltage regulation. Such applicable strategy would surpass the problem due to unbalancing of reactive and active
power [7].
Voltage deficiencies identification at the side of energy supply could be used in their proactive technique and the same method could reduce the
complaint numbers at the demand side. Such a solution is able to decrease the complaint numbers would fail to minimize the occurrence of these
events, therefore, the self-healing approach is able to be applied at this juncture to solve the problem. Forecasting and predictive technologies in
smart power grids make use of algorithms that are able to provide effectively protected systems [8].
Smart grids in the Demand Side Management have to be explored and a variety of load shifting methodologies could be implemented for
effective provision of energy management. Demand-side smart loads are able to be integrated with the exchange data and cellular networks
while making use of the advanced infrastructure in metering [9].
Generally, there are two major problems that relate to communication and power networks highlighted as two-way communication and voltage
regulation structure for overall effective renewable sources of energy distribution. Voltage regulations could be accomplished by substation
5

shunt and load tap charging agents. Synchronization and latency issues in advance communication systems could be monitored by AMI for
increase communication efficiency and response back during troubleshooting [10].
Taking note of the concept in the DSM, the distributed energy storage and distributed energy generation could be recognizable in that they are
major facilitators in the deployment of smart grids. This is due to the problems that come with the integrated renewable sources of energy which
are minimized during dispatch and the DS, distributed energy sources, as well as the DG, distributed energy generation, being incorporated into
innovative DSM methods and the electricity network demand side in a simultaneous implementation [12]. A combination of DS, DG and DSM
techniques would lead to a system that has diverse generation sources that supply energy across the smart grid to bigger sets of uses on the
demand side [13]. The energy efficiency would possibly improve as well as the local generation and load control. The DSM, therefore, could be
referred to as the variety of initiative time modification that is intended to develop the demand magnitude as well as the patterns [13]. Advance
mechanisms can be introduced for the purpose of encouraging the active participation of the demand-side in the process of active optimization.
Hence, users of the demand –side get supplied with control devices, widely known as smart meters. These meters are able to manage the demand
of energy and facilitate the communication with the supply-side [14].
Demand Response
The response that is faster is provided by the variety of (DR) Demand Responses. Specifically, a broadcasting signal such as the transmission or
distribution system operator (TSO/DSO). The signal may be able to contain a command or a price for the shifting/shedding. It is not important
6

for the deadline to be instantaneous. The signal could refer to an instance at 12 noon as the there is a possibility of the grid emergency being
anticipated [15]. There is an assumption by the Classical Direct Load Control to laud fully when under control in they do as they are told [15].
Every intelligence is supposed to be in the controller that ideally makes use of the load model in the production of reasonable decisions. A use of
a stochastic model that are state-space in loads together with the simulation of urban power systems. A technology of this nature has shown both
transport and cost losses. Open ADR is an example of a modern system in the demand response automation. It is developed with the use of a
leading DR research group known as the Demand Response Research Centre. OpenADR has an open-source and open specification reference for
a distributed implementation. The DR orientation of the client-server infrastructure is published I the model of the subscriber. The main
components include [16];
 DRAS that has the site of the consumer clients.
 Demand Response Automation server.
 The communication infrastructure is the internet.
The side of the client is usually a library for communication that is made use of by control manufacturers in making the product capable of
OpenADR. Programs are available for client DR subscription such as the demand bidding and the critical peak pricing with the DRAS being
used as the financial incentives that are connected for reaction to these events [17].
7

The system is nearly an open-loop in that neither the online feedback nor the load models being used. Ripple control is combined using the wide
area system in phasor measurement that uses GPS timestamps. The timestamps are distributed using current/voltage measurement equipment
[18]. An addition of results from the loop feedback is wide in the system energy control a 10% assumption in the load control 1Broadline
powerline communicates a nighttime in the storage of electric stoves [19]. An analysis of the DSM business impacts produces models that are
used in analyzing business models having different simulated electric devices. These are the shiftable loads, loads having storage as well as the
real electric storages [19].
This paper proposes which DSM method, B2G or C2G, that could be used in a smart grid power system for the purpose of optimization with the
attempt to facilitate a reduction of the monetary expense time period during analysis by production and storage of energy rather than the
purchase of their needs of energy from the smart grid. Taking into consideration the user selfish nature, a theoretical approach would particularly
suit a calculated optimum storage and production strategies. For the above-mentioned reason, the modelling of the B2G model is recommended
[20].
Literature Review
A recent source, [21],led to the development of real-time pricing schemes with the aim of reducing the ratio in the peak-to-peak average lauding
through the management of the response demand in systems of the smart grid system. A 2-stage problem in optimization is proposed and later
8

solved. The author comes up with a developed stochastic model that schedule local area networks with the aims being a minimization of cost as
well as PAR minimization.
Studies in [22] present formulation in linear programming that are intended to minimize the cost of energy through direct control of the energy.
A robust approach to the optimization presents an adjustment hourly level of the load for a specific consumer responding to the hourly price of
the electricity. These renewable energies uncertainties, however, were not added to this source. Hence, schemes used in controls were not readily
optimized and be applied to the smart grid scenario that constituted of renewable energies for significant portions of resources of power.
Considering an improvement in energy efficiency, [23] says that a start would be obtaining insight and information for the process to be
involved. Every consumer site would practically have hidden problems that lead to wastage of energy; these include the compressed leakages of
air, dirty filters, misconfigured controls and broken equipment. This trivial problem could be overlooked unless there exists a tool that would be
used in analyzing the efficiency of energy as shown in this source. The parts of information energy that exist include [24];
 Infrastructure for data acquisition. These are the data loggers, sensor networks, modems and gateways.
 A server application possessing a database, analysis algorithms, calculation, reporting and alarming.
 The interface of the users that configures and visualizes.
Classical calculations that exist include [25];
 Comparison of the baseline to the peak loads. High baselines might stem due to an old equipment or a standby power.
9

 Weekly time series comparison. Often ventilation and lighting accidents run through the weekend and the night [1].
 Benchmarks. The performances are compared to others, especially this is critical in multisite users like the supermarket chains.
 Process correlations. The consumption of energy could be checked to strongly correlate with the temperatures on the outside or the gains
in solar.
Other than these static figures inefficiencies, an added dynamics are evident in such systems. Experienced managers of these facilities or use of
smart algorithms make it possible for consumption pattern interpretation and find ways for reducing the peak loads [27]. If the supply of energy
leads to contracted penalized peaks, making it a valuable result. An occurrence of changes in logistics may not help, only equipment automation
is suggested to be a requirement [27].
Energy Controllers
If an equipment operation requires driven consumption adjustment, there could be an implantation of an energy controller. These types of
devices are mainly located in the energy metre to make it possible for the trend of energy consumption to be monitored as stated in [28].
Additionally, the monitored trend indicates unwanted levels, the equipment is switched off by the controller based on the set priorities as well as
other rules. The orientation was depicted as shown below;
10

Another source comes in to mention the configuration of the energy controller to be a task that is very complex. To add on this, the complexity
would increase once the consumers begin to be added and removed. [29] goes on to state that the chosen rules would determine the stability. An
example simple enough for determining the priority level was a dependence on the trend in consumption. A better presentation was set as shown
in the graphs below;
11

12

In these graphs, the author depicts the measurement period that has a country-dependent line time of about 15 or 30 minutes that represent the
smallest period of time for billing reasons. Every period, zero is the starting point and later moves monotonously upwards. The bottom part of
the curve shows where most of the power is consumed that is steeper in the curve of energy consumption. Once the trajectory of consumption of
power crosses either of the upper lines of the threshold, some classes or groups of consumers get switched off. The above graphs show three
devices that are classed. Ci represents the most important devices, c3 is the least important classes [30]. The trajectory begins when it is steep as
every device is turned on initially. The graph exhibits steeper nature than the ideal curve that is dashed hence it goes to cross c3 off turning it off
13

that is the least important for first measures to make the curve flatten. Next is c2 that gets crossed and turned off thereby making the devices
categorized as c1 to be the ones on. The resultant flat line curve would then cross the c1 line but doesn’t turn it off as these devices are left on
anyway. When the lines cross c2 again the devices get turned on and so would the c3 devices ones their line gets crossed. The development of
such a system would make it possible for the consumption energy goal to be attained [31].
Distributed Spinning Reserve
[32] proposes a distributed spinning reserve that is aimed at supporting traditional ancillary service providers in imitation of their behaviour. The
investigation as seen on the demand side meant that these loads could be increased or reduced at instances that the grid frequencies rose or
dropped. There are two implemented schemes which are the Integral Resource Optimization Network that is discussed in [33] and the grid-
friendly controller discussed in [33]. These schemes help in measuring frequency and reacting to it. However, the IRON differs from the GFC
due to the added communication interface allowing cooperate algorithms.
[35] adds on the subject to discuss the add-on feature of the communication fairness capabilities. Taking an example of various devices shedding
their load and one of these devices having insufficient shedding. The existing communication makes it possible for arranging the devices to all
have their turn. Developing this coordination allows stability in the network. A community that could have this autonomy and distributed
controllers facilitating communication would react to the problems of the grid similar to the manner that a perfect unstable recipe. The users
14

would be forced to do it in an orderly manner for avoiding a strong reaction. Such a mechanism distributes control in a classical manner. It
requires the frequency to drop to allow the controllers to react. Also, it requires restoring to prevent regular fixing.
[14]produces a version that is more sophisticated going into the controls that is more-predictive. Devices that service load models of themselves
might be able to predict how long or much the load could be shed until the shedding is topped for process reasons. The models of the load are the
steps followed towards stability. They provide answers to questions of the reaction strength to an anticipated problem needed as well as the
person that can provide it.
Demand Shifting
The load models could be used in the demand needs for shifting to other periods. In case other factors such as weather are predicted, an
emergency in the grid is set in the next day at around 1730. Therefore, [12]produces a research that makes it possible for intelligent users to plan
ahead. This could be done only if the process accepts the planning that may be later or earlier. Some of the examples include; stock production
or precooling. Shifting is then to be done on processes that typically are categorized in one of the lists below;
 Processes in inert diffusion such as irrigation and ventilation.
 Processes that are thermally inert such as cooling and heating,
 Logistics such as dependencies, schedules and lunch-breaks.
 Mass transport such as conveyor belts and pumps having tanks.
15

Load shifting to later points in time is easily done. In instances of critical time, the load has to be shed with the same process being made to catch
up at later periods. However, the quality of the process that is suggested by ################# is not that much guaranteed. If there misses
being proper stocking of products or in instances of the tank being almost empty, the process may get into trouble when it attempts to shed. The
author, therefore, recommends the peak be moved before the time for shedding and preparation to be important. In this model, loads are
required. They avail an anticipated duration of things to be switched off, the duration required for filling the virtual storage as well as the cost.
The demand shifting virtual storage can be enhanced by special means. The researcher goes on to show an instance of phase change in a material
building with an electric heater to increase the low thermal structure of inertia.
Virtual Storage loads Power plants
In a community, Virtual Power Plants are typical small generation units that are depicted as one source having renewable sources of energy
appearing as one grid manage power plant as discussed in [35]. A distributed equipment typically requires a controlled from centralized
management and dispatch node with modern SCADA standards used for integration for the individual parts. Special cases arise in such load
parts. These loads do not generate, they only are virtual storages through load shifting. An aggregate of numerous of these types of loads leads to
the development of sizes capable of participating on the power markets as well as competing with traditional storage of electricity. Aggregators
typically use proprietary technologies in doing this, however, IEC makes a better candidate in enabling the interoperability.
16

[35] mentions that the most important point of VPP is the crucial time in load-based virtual power plants storage, a guaranteed availability. If the
operators of the grids request certain amounts of power regulation, the researcher states that this could be delivered. However, numerous loads
behave in a stochastic manner. A process for customers might in that instance not be interruptible or the virtual storage could be empty. In this
model, once more, reliable load models would be advisable for necessary making the VPP operator possess promises that are available [37].
Load Management Communication Protocol
[37]Researched standard series of IEC used in the automation of substation. The source contains a scheme for addressing IEDs or the DERs as
well as their functions and properties. A substation based on the XML is configured in languages, protocols of communication for preferred
efforts together with real-time transport and so much more. The source states that modern manufacturer’s implement the technology in the latest
products produced in power engineerings such as distributed automation nodes or the diagnostics of devices and the grid measurements. The
model is best implemented in VPPs and gradually makes way down to the distributed level [40].
There have been several groups that publish the energy service standards in the home. One of the groups recommended by [38] is the Zigbee
Smart Energy Profile. The research provides an extension of the description of the document for the large numbers of meter reading services,
security information, online pricing as well as load control. A less descriptive but also interesting proposal would come from the American
Society of Heating, Refrigerating and Air-Conditioning Engineers. There is a defined machine state abstract representing loads without or with
load shedding properties.
17

Research Design
a) Research Question
This paper proposes which DSM method, B2G or C2G that could be used in a smart grid power system for the purpose of optimization with the
attempt to facilitate a reduction of the monetary expense time period during analysis by production and storage of energy rather than the
purchase of their needs of energy from the smart grid [22]. Taking into consideration the user selfish nature, a theoretical approach would
particularly suit a calculated optimum storage and production strategies. For the above-mentioned reason, the modelling of the B2G model is
recommended.
b) Related work
In Consumer-to-Grid (C2G) model systems, the energy being consumed is set to be visualized with the use of a system that communicates back
and forth relaying the information of the consumer [40]. The method tries to engage the consumers present in the household sectors [24]. This
method was chosen due to its ability to reduce the total consumption of electricity averagely to about 7% in one year. Consumers have been
using this technology for some time but its popularity has reduced over time. In this mart grid methodology, external information is required
such the availability of grid congestion and renewable resources in which the signals are received in a continuous manner trying to make the
consumer habit and comfort not to be influenced [25]. The effect was to try demonstrating and analyzing the potential response to demand for
18

the different energy sectors, from home buildings to industrial buildings. Consumers, however, grew to dislike this procedure that monitors
every utilization of energy [26].
Concerning the Building-to-Grid methodology, there is an intended development that would integrate the feedback in energy that would
facilitate automatic responses as well as the user access being independent. Such a feature is aimed at facilitating predictability for every single
user than clustering them as in B2G. Also, the users are able to receive ancillary services that include the control of frequency and control of
voltage [27]. Using B2G involves its implementation to 10 buildings that are selected and the 350kW consumption should be reduced during
their peak times after the installation is complete and working [45]. The load in this model is managed for possible testing of their various
characteristics with an investigation on the advantages, its ability and drawbacks in decentralized and centralized private feedback systems. The
flexibility potential of the consumed energy in the various sectors is assessed to come up with better systems that can manage energy in using
smart grids [28].
c) Design
The selected model for this study is the B2G model in a correlational research for the deep working of the model. Choosing this type of model
was due to its ability to makes use of integration in building system automation in intelligent community software agent cooperation [29]. The
operator of the smart grid interacts with the community negotiation in consumption patterns or asks for support in the emergency case. Its aspect
is therefore modelled to enable the agent to locally access a world model. These voluntary consumers are recommended to be at least 10
19

consumers who may be or may not be in the same building. The model is developed to be having a learning process for the consumers thereby
has the ability to note real-time energy consumption [30]. Using this feature, an agent is enabled to plan ahead as it identifies the typical patterns
of usage, equipment’s time constants and building. The model can then test for an alternative behaviour in advance. The figure below shows how
[31];
Big utility customers were clustered in groups in accordance with the certain rules in statistics like the isolation of demographic differences.
Also, control groups had to be created, consideration of social spread and geography. Every proband then got fitted with smart meters for the
purpose of obtaining fine-grained consumption of energy. As in the figure below [32];
20

Strengths and Weaknesses
Considering the building-to-grid project, it brings back a part that was to be forgotten [33]. The human factor in that the engineers consider
humans and their behaviour as a stochastic or disturbance variable. In energy systems, humans can produce remarkable local intelligence if a
proper inclusion is performed in the required manner. On the other side, technology is bound to be doomed if the users involved do not
understand or like it [34].
It was seen important to learn the functioning of this model during its runtime since it is impossible in the manual configuration of every field
trial building. An intended target B2G is meant to be marked and physically have an oriented demand response [35].
Automatic load shifting was seen to be possible when the consumer systems were also automatic with the external parameters considered as well
as the comfort range. Form these, the acts of communication interface existing between smart grid system and the consumer would be able to
fulfil its smart grid objectives [36].
21

Design Questions
a) Experimental designs are designs that have control over the variables that are involved in the experimentation thereby allowing the
researchers to conclude whether the variables can affect the experimentation. While correlational designs are designs that are developed
to simply have no control over any variable in its experimentation hence the designer has to determine from scratch which factors would
affect the design and which would not.
Experimental designs are most preferred when drawing conclusions that involve establishing or intervening possibilities of influence by
some factors. Also, these designs are straightforward, basic and efficient other than being repeatable for future checking and verification.
However, they are affected by human error with forceful required adherence to ethical standards. They also fail to specify the reason for
an outcome. Correlational designs have easier methods of accomplishment that are less rigorous experimentally since there are reduced
control groups with an existence of an independent variable viable for manipulation. The design also allows the calculation of strengths
in relationships between variables. More so, the design has a pointer that is used for more detailed researches. However, the weakness is
the existence of correlating relationship or the relationship may not be there.
b) Some of the ethical issues that are to be considered include;
Protection of research participants. The rights and welfare of participants that volunteer in a research are important in that there may
erupt series of scandals concerning research in social sciences. No matter the outcome of the research, participants should not be
reprimanded. It is therefore critical for researchers to try protecting their participants.
22

Participants in researchers should also be respected. The study on how people may think, in the conduct of their actions with recorded
observation takes a lot form the volunteers. Anything new would be recorded as well as anything that is repeated in numerous sequences.
Such scrutiny can only be encouraged when the participants are respected. Respect could be attained by the researchers not posting
information that may tarnish the image of its participant.
Participants also have to be informed of the potential benefits of the research both personally and generally to the society. In addition, the
researcher has to review the risks involved thereby making the participant be aware of the results of the agreement. A review of the
research objectives makes the participants know when the contract may be breached and prepares one for the ultimate experimentation.
c) Groups X vs Y is more effective in performance when compared to all other groups followed by group Y vs Z and lastly group X vs Z. it
is also seen that X is the greater influence In performance.
23

References:
[1] J. C. B. H. S. D. R. Sabina, Industrial Internet of Things: Cybermanufacturing Systems, Sydney: Springer, 2016.
[2] P. Palensky and D. Dietrich, "Demand Side Management: Demand Response, Intelligent Energy Systems, and Smart Loads," IEEE Transactions on
Industrial Informatics, vol. 7, no. 3, pp. 381-388, 2011.
[3] C. Y. M. I. M. R. A. Richard, Artificial Intelligence Applications and Innovations: 11th IFIP WG 12.5 International Conference, AIAI 2015, Bayonne,
France, September 14-17, 2015, Proceedings, Sydney: Springer, 2015.
[4] H. Heinrich, Photovoltaics: System Design and Practice, Sydney: John Wiley & Sons, 2012.
[5] T. Jacopo, Peak Energy Demand and Demand Side Response, Sydney: Routledge, 2015.
[6] C. Brebbia, M. Marinov, A. Andrea and C. Safta, Energy and Sustainability IV, Sydney: WIT Press, 2013.
[7] S. Michelle, A. o. S. National and a. M. Engineering, Mathematical Sciences Research Challenges for the Next-Generation Electric Grid: Summary of a
Workshop, Sydney: National Academies Press, 2015.
[8] G. Yogi and K. Frank, Handbook of Energy Efficiency and Renewable Energy, Sydney: CRC Press, 2007.
[9] M. Daphne, H. Peter, L. Victor and B. Richard, Smart Grid Applications and Developments, Sydney: Springer, 2014.
[10] T. Mini and D. John, Power System SCADA and Smart Grids, Sydney: CRC Press, 2017.
[11] B. Stuart, Smart Grids: Advanced Technologies and Solutions, Second Edition, Sydney: Taylor & Francis, CRC Press, 2017.
[12] G. Clark, The Smart Grid: Enabling Energy Efficiency and Demand Response, Sydney: The Fairmont Press, Inc., 2009.
[13] J. Lawrence, Renewable Energy Integration: Practical Management of Variability, Uncertainty, and Flexibility in Power Grids, Sydney: Academic Press,
24

2014.
[14] C. Cecati, C. Citro and P. Siano, "Combined Operations of Renewable Energy Systems and Responsive Demand in a Smart Grid," IEEE Transactions on
Sustainable Energy, vol. 2, no. 4, pp. 468-476, 2011.
[15] L. Nguyen-Thinh, D. Tien, T. Ngoc and A. Hoai, Advanced Computational Methods for Knowledge Engineering: Proceedings of the 5th International
Conference on Computer Science, Applied Mathematics and Applications, ICCSAMA 2017, Sydney: Springer, 2017.
[16] A. Almeida and R. Arthur, Demand-Side Management and Electricity End-Use Efficiency, Sydney: Springer Science & Business Media, 2012.
[17] S. Ruiz-Romero, A. Colmenar-Santos, F. Mur-Pérez and A. López-Rey, "Integration of distributed generation in the power distribution network: The
need for smart grid control systems, communication and equipment for a smart city — Use cases," Renewable and Sustainable Energy Reviews, vol. 38,
no. 1, pp. 223-234, 2014.
[18] D.-F. Bert, P. Boris, R. Christian, U. Dirk, S. Jens-Peter, S. Miranda and Z. Thomas, Balancing Renewable Electricity: Energy Storage, Demand Side
Management, and Network Extension from an Interdisciplinary Perspective, Sydney: Springer Science & Business Media, 2012.
[19] B. Leonard and T. Olivier, Complex, Intelligent, and Software Intensive Systems: Proceedings of the 11th International Conference on Complex,
Intelligent, and Software Intensive Systems (CISIS-2017), Sydney: Springer, 2017.
[20] I. Krzysztof, Smart Grid Infrastructure & Networking, Melbourne: McGraw Hill Professional, 2012.
[21] S. IEEE, 2017 Smart Grid Conference (SGC), Melbourne: IEEE, 2017.
[22] "Voltage regulation in distribution feeders with high DG penetration: From traditional to smart," in IEEE PES General Meeting: The Electrification of
Transportation and the Grid of the Future, vol. 1, no. 3, pp. 1-7, 2018.
[23] M. Juan, C. Antonio, M. Henrik, P. Pierre and Z. Marco, Integrating Renewables in Electricity Markets: Operational Problems, Melbourne: Springer
Science & Business Media, 2013.
[24] L. Kwang and E.-S. Mohamed, Modern Heuristic Optimization Techniques: Theory and Applications to Power Systems, Melbourne: John Wiley & Sons,
25