Demand-Response: Let the Devices Take our Decisions

Guillaume Guerard, Bastien Pichon and Zeinab Nehai

eonard de Vinci P

ole Universitaire, Research Center, 92916 Paris La D

efense, France

Keywords:

Demand-Side Management, Demand-response, Smart Grid, Automaton, Game Theory.

Abstract:

The hierarchical, centrally controlled energy grid is ill-suited to the third digital revolution. The electric power

industry is undergoing rapid change. The energy transition move from the current energy system using non-

renewable resources to a smart grid, including distributed resources and home automation. Now, the demand is

ﬂexible and can be managed, it is called Demand-Side-Management (DSM). It encompasses different domains

of reducing consumption, it can be both a physical standpoint than digital. In this paper, after a quick state of

an art on DSM, we will focus on the digital way. The main idea is to create consumption’s schemes, thanks to

home automation in order to ﬁnd the best way to consume.

1 INTRODUCTION

The creation of the Smart Grid has been posed as one

of the greatest challenges of this century, as countries

face dwindling non-renewable energy sources and the

adverse effects of climate change due to carbon emis-

sions

The vision of a Smart Grid includes technologies

that enable the efﬁcient integration of new sources

of energy. It will smooth demand by allowing con-

sumers to better manage how electricity is used,

stored, and delivered. However, the balance between

demand and production is not an easy task. Both sup-

ply and demand levels can change rapidly due to out-

ages, sudden load change or volatile renewable en-

ergy sources.

The term Demand-Side Management, introduced

in 80s by the Electric Power Research Institute, refers

to all strategies that can reduce the consumption and

peak demand (e.g., in (Ruiz et al., 2009; Centolella,

2010)). The trend started in the 70s, and became

rapidly a government project.

The DSM encompasses all means possible in re-

ducing energy in a house or a building: a top wall in-

sulation prevents heat dissipation, so a thermal energy

loss; light sensors avoid waste lighting; smart devices

reduce power consumption; photo-voltaic panels on

the roof produce electricity for a home.

One of recent industrial developments concern the

concept of the smart meter which aims to manage the

US DOE, Grid 2030: A national vision for electricity’s

second 100 years, 2003

devices in the home to minimise inefﬁciencies in us-

age and maximise the user’s savings. Smart meters

also aim to interact with the grid in order to help re-

duce peaks in demand and keep up with variable en-

ergy generators or batteries

. This approach has been

shown to be effective for small pool sizes of indus-

trial and commercial consumers (Hammerstrom et al.,

2007).

While such DSM techniques have been shown to

bring about signiﬁcant improvements on a small num-

ber of houses, it is unclear how such technologies will

scale when smart meters with millions of homes or

buildings nationwide. In particular, the centralized

management of even thousands of smart meters is

likely to be a complex task that may require intrud-

ing upon user’s privacy to cater for all homes.

The fact that increasingly more and more features

of the home are likely to be electriﬁed in the future

(Galvin and Power, 2009), means that more signiﬁ-

cant peaks may be created due to the reactive behavior

of the smart devices. Nonetheless, without some form

of coordination, the population of smart grid entities

may end up with overly-homogeneous optimized con-

sumption patterns that may generate signiﬁcant peaks

in demand in the grid.

In a collaborative system such as the smart grid,

fairness and stability is important when taking a deci-

sion. If, for example, a number of devices can shed

their load, and already one of them shedding is suf-

ﬁcient, it might always be the quickest that wins the

DECC Smarter Grids report, The Opportu-

nity,December 2009

Guérard, G., Pichon, B. and Nehai, Z.

Demand-Response: Let the Devices Take our Decisions.

DOI: 10.5220/0006196601190126

In Proceedings of the 6th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2017), pages 119-126

ISBN: 978-989-758-241-7

119

monetary incentive. With communication, it can be

arranged that all of them have their turn. Without

communication, all of them react to grid problems in

the same manner (Palensky and Dietrich, 2011). It is

the perfect recipe for instabilities.

There is a need to deploy innovative models and

algorithms that capture the following characteris-

tics of the emerging smart grid: communication in

a heterogeneous system, distributed operation, low-

complexity distributed algorithms (Saad et al., 2012).

This paper provides a classiﬁcation of DSM pro-

grams and a categorization of devices for home au-

tomation management. It also exposes a descrip-

tion of a model that would optimize the Demand-

Response, a DSM programs which focuses on a digi-

tal aspect. This model is generic and is applicable to

any network.

First of all, in section 2, we introduce the notion of

Demand-Side Management, a classiﬁcation about the

intrinsic concepts. Then we expose a solution in sec-

tion 3 about the management of smart devices through

Demand-Response strategies. The section 4 describes

how to build consumption’s schemes, thanks to home

automation and generic modeling. The section 5

shows a game theoretic approach to avoid peak de-

mand and optimize each consumer’s comfort.

2 DEMAND SIDE MANAGEMENT

DSM is a fuzzy concept which has various deﬁnitions.

DSM programs usually refer to one or both of the fol-

lowing design objectives: reducing consumption and

shifting consumption. DSM can be as well an energy

reduction by using insulating materials for the home,

or using devices controlled numerically.

2.1 Introduction to DSM

In (Saad et al., 2012), a survey about DSM pro-

grams, the authors deﬁne DSM as programs that at-

tempt to make the users more energy-efﬁcient on a

longer time-scale. They also deﬁne the term Demand-

Response as programs that utility companies use to

encourage the grid users to dynamically change their

electricity load so as to have a short-term reduction in

energy consumption.

In other words, the goal of the DSM is to en-

courage the consumer to use less energy during peak

hours, or to move the time of energy use to off-peak

times such as nighttime. Peak demand management

does not necessarily decrease total energy consump-

tion, but could be expected to reduce the need for in-

vestments in networks and/or power plants for meet-

ing peak demands.

First DSM programs are based on energy efﬁ-

ciency measures. They include all permanent changes

on equipment or improvements in the physical prop-

erties of the system (Boshell and Veloza, 2008).

Numeric DSM programs are based on two

schemes: direct load control and smart pricing. They

focused on the interactions between producers and

each individual end-user. Direct control enables to

control the appliances inside a building, smart pricing

provides monetary incentives for the users.

The deﬁnition varies, but in all cases, the beneﬁts

of DSM programs are for all actors. An overall elec-

tricity price reduction is expected because of a more

efﬁcient using of the infrastructure. The smart grid fa-

vors using local renewable energies or batteries than

to activate coal plants. Moreover, DSM programs can

increase short-term capacity using market-based pro-

grams, which in turn, results in an avoided or deferred

capacity costs.

The generation cost increases exponentially near

maximum generation capacity. A small reduction in

demand will result in a big reduction in generation

cost and, in turn, a reduction in electricity price, that

affect all market participants. All of the avoided or de-

ferred costs will be reﬂected in the price of electricity

for all electricity consumers.

By having a well-designed DSM program, users

also have the opportunity to help in reducing the risk

of outages. Simultaneously and as a consequence,

they reduce their own risk of being exposed to forced

outages and electricity interruption. On the other

hand, the operator will have more options and re-

sources to maintain system reliability, thus reducing

forced outages and their consequences (Goel et al.,

2006).

Rather, a bad DSM program creates a rebound ef-

fect (or payback), is typically not saved and maybe

even a new peak is generated. In this ﬁgure, EE means

Energy Efﬁciency and DR means Demand-Response.

We will deﬁne those two terms in the following sub-

section.

2.2 Classiﬁcation

This section is dedicated to classifying some pro-

grams by a set containing similar programs. The Fig-

ure 1 shows the classiﬁcation.

We usually use devices or appliances at work and

at home. Thus, social behaviors have a big impact

on consumption. The Change Management (CM) is a

set of incentive programs and social programs in or-

der to create responsible behaviors, like sorting the

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

120

Figure 1: Classiﬁcation of DSM types.

wastes in the late 90s. People need some information,

knowledge and teaching about home automation, sus-

tainable energy, energy management’s dashboard or

any other current of future technologies which will

interact in the smart grid. By the way, CM has to give

awareness about the two other edges of DSM: Sus-

tainable Energy (SE) and Demand-Response (DR).

The SE includes all permanent changes in equip-

ment or improvements in the physical properties of

the system (Boshell and Veloza, 2008). By deﬁni-

tion, Energy Efﬁciency (EE) is the goal to reduce the

amount of energy required to provide products and

services. Using less power to perform the same tasks.

In simple words, SE involves a permanent reduc-

tion of demand by using more efﬁcient load-intensive

appliances such as water heaters, refrigerators, or

washing machines. For example, insulating a home

allows a building to use less heating and cooling en-

ergy to achieve and maintain a comfortable tempera-

ture. Installing ﬂuorescent lights, LED lights or nat-

ural skylights reduces the amount of energy required

to attain the same level of illumination compared with

using traditional incandescent light bulbs.

Improvements in the SE are generally achieved

by adopting a more efﬁcient technology or produc-

tion process or by application of commonly accepted

methods to reduce energy losses.

The last but not least edge of DSM is the DR, it

allows to know how electricity consumers can be re-

sponsive. DR programs follows, including classical

incentive programs, new market-based and dynamic

pricing scenarios, besides potential cost savings and

beneﬁts related to different market components, it

performs digital control of consumption.

2.3 About Demand-Response

According to the Federal Energy Regulatory Com-

mission, DR is deﬁned as:

The changes in electricity usage by end-use

customers from their normal consumption pat-

terns in response to changes in the price of

electricity over time, or to incentive payments

designed to induce lower electricity use at

times of high wholesale market prices or when

system reliability is jeopardized.

To resume, DR includes all intentional modiﬁca-

tions to consumption patterns of electricity to induce

customers that are intended to alter the timing, level

of instantaneous demand, or the total electricity con-

sumption (Albadi and El-Saadany, 2007).

We notice two categories of devices that you can

apply DSM: those that includes DR programs and

those that don’t. Examples of the latter include light-

ing, entertainment devices, phone charging and com-

puter usage. Those devices interact with the grid

through automatic process thanks to sensors or inte-

grated circuit. This kind of process cannot be quali-

ﬁed as ”smart”, i.e. not ﬂexible and not adaptive.

Demand-Response: Let the Devices Take our Decisions

121

DR programs can be classiﬁed into two main

categories: Incentive-Based Programs (IBP) and

Price-Based Programs (PBP) (Albadi and El-Saadany,

2008; Cappers et al., 2010).

In IBP, participating customers receive participa-

tion payments, usually as a bill credit or discount rate.

Those following programs are part of IBP: direct load

control, curtaible rates, emergency demand response,

capacity market program, demand bidding program

(Ramchurn et al., 2011).

PBP programs are based on dynamic pricing rates

in which electricity tariffs are not ﬂat; the rates ﬂuc-

tuate following the real time cost of electricity. These

rates include the Time of Use (TOU) rate, Critical

Peak Pricing (CPP), Extreme Day Pricing and Real

Time Pricing (RTP) (Rahman et al., 1993; Ramchurn

et al., 2011). Many studies on PBP programs offer an

overview of the rebound effect.

The TOU pricing simply biases the real price of

electricity in order to incentive users who typically

aim to maximize their savings, to shift their loads

to off-peak periods (i.e., when aggregate demand is

lower). However, where the price of electricity at

night is cheaper than during the day, the TOU pric-

ing has been observed to create signiﬁcant additional

peaks in demand as soon as the off-peak period is

reached (Ramchurn et al., 2011; Strbac, 2008).

There exists a lot of alternatives, hybrid programs

such as Smart Pricing, where users are encouraged to

individually and voluntarily manage their loads, e.g.,

by reducing their consumption at peak hours (Cen-

tolella, 2010; Herter, 2007). Some scheduling pro-

grams using CPP, TOU pricing, and RTP are among

the popular options.

DR includes automatic process and decision mak-

ing process. In this paper, we present a decision mak-

ing model to manage home automation in response

to an RTP based on produced energy price, not on a

supply/demand energy market price.

3 PROCESS OF OUR MODEL

A smart grid must allow customers to make informed

decisions about their energy consumption, adjusting

both the timing and quantity of their electricity use.

The process of our model is as follows:

Step 1, Data Update: at the beginning of a new iter-

ation, sensors and automata update data.

Step 2, Consumption’s schemes (Figure 2):

through a knapsack problem and thanks to

automaton, consumption’s schemes are built

(Section 4). They represent all the consumption

possibilities in a smart house.

Step 3, Game for Demand-Response (Figure 3):

a game between each consumer and producer

is created. Strategies depend on consumption’s

schemes and producers’ response. The best

economic choice in the game is chosen (Section

5).

Step 4, Decision: following to the previous decision,

the smart grid computes how energy is routed

across the grid. In function of the result, the ﬁnal

decision is taken or a feedback adjusts the game.

The process of the whole model is presented in

(Ahat et al., 2013). This paper improves the local and

microgrid management.

4 CONSUMPTION’S

STRATEGIES

Technologies are available, and more are under devel-

opment, to automate the process of DR. Such tech-

nologies detect the need for load shedding, commu-

nicate the demand to participating users, schedule

load shedding, and verify compliance with demand-

response programs. GridWise and EnergyWeb are

two major federal initiatives in the United States to

develop these technologies. Universities and private

industry are also doing research and development in

this ﬁeld.

In this section, we provide our approach through

automation and a set of consumption’s schemes. Each

prosumer build its set on each possibility of consump-

tion of the local devices. Then, the set is sent to the

microgrid. In function of how the grid will react, the

best strategy is chosen for each prosumer through a

game. Figures 2 and 3 show an overview of the pro-

cess, see references in section 3 for more information

about the process.

4.1 DR Programs

Let us remind you the deﬁnition of DR programs.

They include all intentional electricity consumption

pattern modiﬁcations by end-use customers that are

intended to alter the timing, level of instantaneous de-

mand, or total electricity consumption (Albadi and El-

Saadany, 2008).

There are four general actions by which a cus-

tomer response can be achieved (QDR, 2006; Sezgen

et al., 2007; Valero et al., 2007):

Reducing Power: customers can reduce their elec-

tricity usage during critical peak periods when

prices are high without changing their consump-

tion pattern during other periods.

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

122

Figure 2: From devices to consumption’s schemes.

Figure 3: From consumption’s schemes to a microgrid game.

Shifting: customers respond to high electricity prices

by shifting some of their peak demand operations

to off-peak periods.

Spinning Reserves: customers use local batteries

and distributed generation combined with loads’

management.

Emergency Cut: when the peak demand is still ac-

tive even after the ﬁrst three actions, an emergency

cut is operated.

Reducing power involves turning down or off spe-

ciﬁc appliances. For example, heating may be turned

down or air conditioning or refrigeration may be

turned up, delaying slightly the draw until a peak in

usage has passed (Sinitsyn et al., 2013). The success

of such programs depends on a suitable pricing sys-

tem for electricity.

In shifting load, the load is shed at the critical time

and the process has to catch up later. For this, load

models are needed. They predict how long devices

can be turned off, how much it takes to ﬁll the storage,

and what it costs (Kupzog and Palensky, 2007). The

strategies to manage loads are complex and need to

communicate with the grid to establish a schedule.

Spinning reserve is done with batteries. It tries to

support the traditional providers of ancillary services

by imitating their behavior. Two implementations of

this scheme are the Integral Resource Optimization

Network (IRON) (Stadler et al., 2005) and the grid-

friendly controller (Cantin et al., 1995). Both mea-

sures the frequency and react to it. The difference is

that IRON has an additional communication interface

that allows cooperative algorithms, i.e. a consensus

among devices.

Energy storage is expected to be a key component

in smart homes, and, thus, it has a strong impact when

Demand-Response: Let the Devices Take our Decisions

123

used with home automation and local renewable ener-

gies. For example, a user may decide to store energy

during off-peak hours and use this stored energy to

schedule its appliances, instead of obtaining this en-

ergy directly from the substation during peak hours.

4.2 Devices Classiﬁcation

There are many ways to classify devices. Some clas-

sify devices according to their usefulness or inner

workings (wet, cold, water heating, etc.) (Hamidi

et al., 2009). We follow some simple rules in our

classiﬁcation (see Figure 4): can the device follow

DR programs? which data are useful for a device?

The classiﬁcation is made in order to categorize all

existing devices and the ones which will be created.

The No DR programs gather devices which can-

not be controlled by a decision making process, but

still can be managed by sensors. The devices may be

Cyclic or Acyclic, and with Sensors.

Cyclic means the devices which have an operating

cycle (i.e. boilers, coffee maker), they have a start, an

operating time deﬁned, and an end. These kind of de-

vices cannot be controlled because their uses are un-

predictable and immediate. So do the acyclic devices

which encompass the most household machines (i.e.

vacuum cleaner, blender, TV, radio, microwave, iron,

modem, telephone, DVD player, printer, oven).

The third one includes devices controlled by sen-

sors, it concerns more generally lighting. The light-

ing control is based on a brightness sensor assembly,

which estimates the amount of light for a room ac-

cording to the natural light received from the outside.

In this model, devices are controlled by automa-

ton in order to optimize their own consumption. A

device is deﬁned by a set of four categories, it picks

one element of each category:

1. the device is cyclic (i.e. washing machine) or

acyclic (i.e freezer)

2. the device has or doesn’t have batteries

3. which data are used: external (i.e. heater in the

room) or internal (i.e internal temperature of the

refrigerator)

4. the device consume, or produce, or both.

Then, a device gets a set of parameters. They are

input values that determine how and when a device

will consume or adjust its consumption following a

DR program. The set includes:

Internal data: they correspond to internal data of the

machine, collected by internal sensors (i.e. tem-

perature inside the fridge).

External data: they correspond to external data of

the machine, in its environment collected by sen-

sors (i.e. the temperature data of a room).

Consumer’s preference: these parameters are the

most important. Because the user should not be

hampered in his daily life, its preferences are the

bounds that devices have to reach (i.e. which tem-

perature the consumer prefers in its bedroom).

Price: consumers, as described in (Kirschen et al.,

2000), consider both current prices and the prices

of one step into the future. To perform shifting,

devices have to know how to schedule their con-

sumption. In our model, we use current prices,

prices of one step into the future and price trend.

The last two are calculated from derivative func-

tion, norms and a pricer.

5 GAME THEORY AND BEST

CONSUMPTION’S

STRATEGIES

The authors in (Mohsenian-Rad et al., 2010) show

that it is better to develop a DR approach that opti-

mizes the properties of the aggregate load of the users.

This is enabled by the deployment of communication

technologies that allow the users to coordinate their

energy usage, when this is beneﬁcial.

The essence of DR revolves around the interac-

tions between various entities with speciﬁc objectives

which are reminiscent of the players’ interactions in

game theory. As Saad et al. said (Saad et al., 2012):

Game theory provides a plethora of tools that

can be applied for pricing and incentive mech-

anisms, scheduling of appliances, and efﬁcient

interconnection of heterogeneous nodes.

In our model, each smart meter sends all the con-

sumption’s schemes, deﬁned by their automaton, of

their devices. A strategy for a consumer is a combi-

nation of a possible scheme of consumption of each

device. Thus, the number of strategies is a combina-

torial set. For example, if the smart building has four

devices with respectively 2, 3, 3, 4 schemes, the num-

ber of strategies is equal to 72.

The price signal to incentivise the agents may de-

fer their demand. Even if an accurate price signal is

provided, the adaptive and autonomous behaviour of

the agents in the system is a key component that can

enable signiﬁcant performance beneﬁts in the smart

grid. It’s important to determine how comfort is more

important to the user relative to price. This is done

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

124

Figure 4: Characterization of a device.

Figure 5: Lipschitz function that manages DR programs.

following the Palenksy et al.’s works (Palensky and

Dietrich, 2011).

Both parameters depend on how much the current

consumption’s curve is far from an ideal consump-

tion’s curve as shown in Figure 5. DR programs like

emergency programs or reducing programs depends

on the value of the k-Lipschitz function.

If the computed load is larger than the previous

one, a DR program is launched. In opposite case,

if the curve is smaller than the previous one, more

devices consume or return to a basic mode of con-

sumption. Different emergency programs trigger de-

pending to how much k in the Lipschitz function is

large. In this way, the global consumption curve ﬂat-

tens over time.

With communication, the shed can be arranged by

the game, i.e. in the same microgrid. Such coordina-

tion also contributes to stability. Imagine a commu-

nity of autonomous, distributed controllers without

communication. All of them reacting to grid prob-

lems increase instabilities. They will do it one after

another to avoid a too strong reaction (Palensky and

Dietrich, 2011).

We don’t argue about the second player is this

paper. This one represents producers, user’s con-

tract and some other properties that depend on gov-

ernment policies about DR. We don’t present an util-

ity function. It depends on the price’s values, user’s

preference, and the feedback function (Gu

erard et al.,

2015). Those works will be shown in a future paper

with the feedback process.

Once a strategy is valid, i.e. can be routed from

the producers to the microgrid, the last one send a sig-

nal to the smart building. This consensus is reached

in the whole grid at the same time. Thus, every con-

sumer and prosumer know how to adapt their behav-

ior for the next step. All the generated data is useful

to compute forecasts, future production scheme and to

be used in data mining and machine learning (Gu

erard

et al., 2015).

6 CONCLUSION

There is a need to deploy new models and algorithms

that can capture the following characteristics of the

emerging smart grid. It is a current and active ﬁeld

that will give birth to many innovations and technolo-

gies. The needs to build an efﬁcient and ﬂexible smart

grid are known, and it becomes an urgent matter while

population and technologies increase drastically.

The presented model provides some simple and

useful tools for a generic model of smart grid. This

decision making tool can be used to test existing or

future technologies in a smart grid design. As any

multi-agent system, this model can be set as wanted.

Examples will be made on different sets of microgrid

and a small smart grid.

REFERENCES

Ahat, M., Amor, S. B., Bui, M., Bui, A., Gu

erard, G., and

Petermann, C. (2013). Smart grid and optimization.

Demand-Response: Let the Devices Take our Decisions

125

American Journal of Operations Research, 3:196–

206.

Albadi, M. H. and El-Saadany, E. (2007). Demand response

in electricity markets: An overview. In 2007 IEEE

power engineering society general meeting.

Albadi, M. H. and El-Saadany, E. (2008). A summary of de-

mand response in electricity markets. Electric power

systems research, 78(11):1989–1996.

Boshell, F. and Veloza, O. (2008). Review of developed de-

mand side management programs including different

concepts and their results. In Transmission and Dis-

tribution Conference and Exposition: Latin America,

2008 IEEE/PES, pages 1–7. IEEE.

Cantin, L., Deschenes, M., and D’amours, M. (1995). Con-

troller for controlling operation of at least one elec-

trical load operating on an ac supply, and a method

thereof. US Patent 5,442,335.

Cappers, P., Goldman, C., and Kathan, D. (2010). De-

mand response in us electricity markets: Empirical

evidence. Energy, 35(4):1526–1535.

Centolella, P. (2010). The integration of price responsive

demand into regional transmission organization (rto)

wholesale power markets and system operations. En-

ergy, 35(4):1568–1574.

Galvin, R. Y. and Power, K. P. (2009). How the micro-

grid revolution will unleash cleaner, greener, and more

abundant energy. Nova Iorque: McGraw-Hill.

Goel, L., Wu, Q., and Wang, P. (2006). Reliability enhance-

ment of a deregulated power system considering de-

mand response. In 2006 IEEE Power Engineering So-

ciety General Meeting, pages 6–pp. IEEE.

erard, G., Amor, S. B., and Bui, A. (2015). A context-

free smart grid model using pretopologic structure.

In Smart Cities and Green ICT Systems (SMART-

GREENS), 2015 International Conference on, pages

1–7. IEEE.

Hamidi, V., Li, F., and Robinson, F. (2009). Demand re-

sponse in the uk’s domestic sector. Electric Power

Systems Research, 79(12):1722–1726.

Hammerstrom, D. J., Brous, J., Chassin, D. P., Horst, G. R.,

Kajfasz, R., Michie, P., Oliver, T. V., Carlon, T. A.,

Eustis, C., Jarvegren, O. M., et al. (2007). Paciﬁc

northwest gridwise testbed demonstration projects;

part ii. grid friendly appliance project. Technical re-

port, Paciﬁc Northwest National Laboratory (PNNL),

Richland, WA (US).

Herter, K. (2007). Residential implementation of critical-

peak pricing of electricity. Energy Policy, 35(4):2121–

2130.

Kirschen, D. S., Strbac, G., Cumperayot, P., and

de Paiva Mendes, D. (2000). Factoring the elastic-

ity of demand in electricity prices. IEEE Transactions

on Power Systems, 15(2):612–617.

Kupzog, F. and Palensky, P. (2007). Wide-area control

system for balance-energy provision by energy con-

sumers. IFAC Proceedings Volumes, 40(22):337–345.

Mohsenian-Rad, A.-H., Wong, V. W., Jatskevich, J.,

Schober, R., and Leon-Garcia, A. (2010). Au-

tonomous demand-side management based on game-

theoretic energy consumption scheduling for the fu-

ture smart grid. IEEE transactions on Smart Grid,

1(3):320–331.

Palensky, P. and Dietrich, D. (2011). Demand side manage-

ment: Demand response, intelligent energy systems,

and smart loads. IEEE transactions on industrial in-

formatics, 7(3):381–388.

QDR, Q. (2006). Beneﬁts of demand response in electricity

markets and recommendations for achieving them. US

Dept. Energy, Washington, DC, USA, Tech. Rep.

Rahman, S. et al. (1993). An efﬁcient load model for

analyzing demand side management impacts. IEEE

Transactions on Power Systems, 8(3):1219–1226.

Ramchurn, S. D., Vytelingum, P., Rogers, A., and Jennings,

N. (2011). Agent-based control for decentralised de-

mand side management in the smart grid. In The 10th

International Conference on Autonomous Agents and

Multiagent Systems-Volume 1, pages 5–12. Interna-

tional Foundation for Autonomous Agents and Mul-

tiagent Systems.

Ruiz, N., Cobelo, I., and Oyarzabal, J. (2009). A direct load

control model for virtual power plant management.

IEEE Transactions on Power Systems, 24(2):959–966.

Saad, W., Han, Z., Poor, H. V., and Basar, T. (2012). Game-

theoretic methods for the smart grid: An overview

of microgrid systems, demand-side management, and

smart grid communications. IEEE Signal Processing

Magazine, 29(5):86–105.

Sezgen, O., Goldman, C., and Krishnarao, P. (2007). Op-

tion value of electricity demand response. Energy,

32(2):108–119.

Sinitsyn, N. A., Kundu, S., and Backhaus, S. (2013).

Safe protocols for generating power pulses with het-

erogeneous populations of thermostatically controlled

loads. Energy Conversion and Management, 67:297–

308.

Stadler, M., Palensky, P., Lorenz, B., Weihs, M., and Roe-

sener, C. (2005). Integral resource optimization net-

works and their techno-economic constraints. na.

Strbac, G. (2008). Demand side management: Beneﬁts and

challenges. Energy policy, 36(12):4419–4426.

Valero, S., Ortiz, M., Senabre, C., Alvarez, C., Franco, F.,

Gabald, A., et al. (2007). Methods for customer and

demand response policies selection in new electricity

markets. IET generation, transmission & distribution,

1(1):104–110.

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

126