See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/277532765

Microgrid's Energy Management Systems: A Survey

Conference Paper · May 2015

DOI: 10.1109/EEM.2015.7216662

CITATIONS READS
13 2,800

3 authors:

I.D. Serna-Suarez G. Ordonez

Industrial University of Santander Industrial University of Santander
6 PUBLICATIONS   20 CITATIONS    94 PUBLICATIONS   236 CITATIONS   

SEE PROFILE SEE PROFILE

Gilberto Carrillo-Caicedo
TecnoGestión
33 PUBLICATIONS   437 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mejoramiento del desempeño de paneles fotovoltaicos a partir de la irrigación forzada de la superficie superior View project

Non-intrusive Load Monitoring View project

All content following this page was uploaded by Gilberto Carrillo-Caicedo on 01 June 2015.

The user has requested enhancement of the downloaded file.

 1

Microgrid’s Energy Management Systems: A Survey

I.D. Serna-Suárez, G. Ordóñez-Plata G. Carrillo-Caicedo
Escuela de Ing. Eléctrica, Electrónica y de Maestría en sistemas energéticos avanzados
Telecomunicaciones Universidad de Santander
Universidad Industrial de Santander Bucaramanga, Colombia
Bucaramanga, Colombia [email protected]
[email protected], [email protected]

Abstract—Microgrids are a new paradigm for energy distribu- explained by the results of microgrid studies, which indicate
tion systems in which generation (from a local energy source or that microgrids are technically and economically efficient –and
storage device) is coordinated to supply local energy needs while profitable.
behaving as a sole system. Benefit maximization in a microgrid is
directly related with demand reactiveness and optimal allocation A close analysis of the study from Inter-American Develop-
of local energy resources. Because those characteristics can be ment Bank (IDB) in [5] shows four problems: generation cost
implemented through an Energy Management System (EMS), is still too high, there is not a common methodology to evalu-
knowing the elements that conforms it and how they interact is ate benefits, intangible benefits are usually reduced to emission
crucial for the development of microgrids. This paper presents mitigation (like social or environmental benefits), and benefits
a caracterization of the EMS and some insight about the actual
state on this subject. This is done to understand the mechanisms are analyzed only in local scenarios. To those problems one
that allow efficient EMS implementations, thus shortening the should add the lack of an "adaptation process of the state-of-
path towards a power distribution microgeneration with higher the-art technologies to local and regional scenarios" [7] which
short-term/mid-term benefits based on microgrid technology. inhibit massification of needed technologies –this is why in [7]
Index Terms—Decentralized optimization, energy management this element is recognized as a relevant goal for researches on
systems, energy storage systems, forecasting, multiagent systems. non-conventional renewable energy sources.
This work will explore one solution to the generation
cost problem by analyzing the Energy Management Systems
I. I NTRODUCTION (EMSs) in a microgrid, which might increase the revenue
Energy potential calculations show that all final demand stream in a microgrid by taking advantage of the benefits that
energy needs can be satisfied with non-conventional renewable they offer.
energy (e.g. from sun, from wind or from biomass) [1]. A full review of energy management systems is unpractical
Moreover, technology advances are reducing manufacturing in the present document, so the first issue addressed is a
and operational costs of technology needed to capture such matter of definition (section II). Next, some details of energy
energy sources. As a consequence, renewable energy has management systems are treated based on the characteristics of
become a key element in the future energy portfolio of any common subsystems for (almost) all energy management sys-
nation. tems such like the forecasting module (section III), the energy
Microgenerators are machines designed to capture energy storage module (section IV) and the operation optimization
from local sources, including non-conventional renewable en- module (section V). This work ends with a discussion and
ergy. The prefix micro indicates that generator capacity is some concluding remarks about energy management systems.
lower than 5 MW. On the other side, a microgrid is a new
paradigm for energy distribution systems in which flexible II. D EFINITIONS
loads and the energy from microgenerators and storage devices As presented in this paper, the concept of Energy Man-
are coordinated to supply local energy needs while behaving agement Systems (EMSs) take into account the definition
as a sole system. of microgrid found in [6], [8]–[12]. From there, one may
The present technology makes affordable microgenera- conclude that a distributed resource island system makes also
tion installations [2] and full microgrid implementations [3], a microgrid [8], [13]. Thus, a microgrid:
[4]. But there is a catch: most microgeneration installations  Includes a local electric power distribution system or a
need regulation aids and most microgrid implementations are portion of such a system.
merely test systems. Thus, a question may arise: what is  Has Distributed Energy Resources (DERs) and loads.
holding back microgrid massification?  Can operate connected to the main grid (grid-connected
Reference [5] focuses in the calculation of the social bene- mode) or independently from it (islanded mode).
fits, and show through the Levelised Cost Of Energy (LCOE)  Is intentional and planned.
approach that in this very moment microgeneration social ben- It worth noticing that DERs may include distributed generation
efits compensates any cost disadvantage. More interestingly, in systems and/or distributed storage systems.
[6] the authors claim that to make microgrids profitable one Formally, an energy management system is «[...] a collec-
must use them to combine several of its benefits. Therefore, tion of control strategies and operational practices, together
the lack of a massive microgrid implementation cannot be with the hardware and software to accomplish the objectives
 2

of energy management.» [14]. This work excludes software the one-day-ahead forecast error, it will be wise to choose the
considerations related to communication problems and all longer scope [22].
hardware considerations. In addition, only high level control With a characterized variability, the next question lies on
strategies and operational practices for a single power distribu- tool selection. In [24], the authors claim that auto regressive
tion system are analyzed. High level controls are related to the mean average (ARMA) models are well suited for predictions
optimization problem statement and solution, whilst low level up to 10 hours in wind velocity estimation, surpassing numer-
controls are related to microgrid’s power electronic interface ical weather prediction methods in the range of 4 to 6 hours
operation. For a time-based classification of control strategies ahead. For a day ahead forecast, artificial neural networks
see [15] and for a hierarchical-based classification of control seem to be a recurrent option [18], [19], [22], [25].
strategies see [16]. Nonetheless, there are some other tools: In reference [22]
To analyze the characteristics of Energy Management Sys- also is used an averaged index over the period of interest to
tems (EMSs) the architecture depicted in Fig. 1 is pro- forecast solar irradiance. Palma-Behnke et al. clarifies that this
posed. Following sections describe the characteristics of the is possible because the altitude of the test system is above
forecasting, storage management system and the operation the 20°; In reference [15] authors predict the load behavior
optimization modules. with the total energy approach; and in reference [21] is used
a Markov process to incorporate the forecast problem into a
reliability analysis.
Weather Local demand &
Forecast generation
conditions
IV. S TORAGE MANAGEMENT
Depending on the available storage devices and local gener-
ation equipment the storage management could be composed
Fuel & Electricity Reserve
Storage status by only a battery or several storage devices, both cases are
prices management treated bellow.

A. Battery management
Optimization This kind of storage is typically a lead-acid battery. Bat-
tery systems are a sensible element in Energy Management
EMS Systems (EMSs) because «how [it] performs over its lifetime
is highly dependent upon how it is operated in real-time
Power output
setpoints during each charge and discharge cycle.» [26]. Battery life
cycle optimization needs to control battery state of charge and
Fig. 1. Typical Energy Management Systems (EMS) architecture. Forecast charge/discharge rates. This is better accomplished if charging
is needed to uncertainty mitigation of input parameters. Adapted from [17], policies are independent from the EMS. Two main approaches
[18]. have appeared to handle this: Rule-based methods and Fuzzy
Logic methods.
III. F ORECAST Chen et al. in [25] show a rule-based method which consists
Energy Management Systems (EMSs) analysis must include on changing the constraints (e.g. maximum charge/discharge
forecasting process because better estimation of some opti- rates) on the optimization problem depending on the actual
mization parameters improves the optimization results [19], state of the battery. A rule-based approach depending on actual
[20]. Most forecasting is done over 24 time slots, each cor- and future (day-ahead) weather conditions is presented in [17].
responding to an hour of a day. This day-ahead forecasting The fuzzy logic approach in [18] use the battery state
is often applied to predict weather conditions, load levels and of charge, electricity price, load demand, renewable energy
market prices. generation, actual solar daily generation and next day solar
More time slots could seem to be a good option to improve availability to derive a charge or discharge signal. On the
optimization results, however, as a consequence of the high other side, in [20] the storage system depends mainly on
variability of some local sources (e.g. wind), forecasts become the charging price: the difference between the maximum
less reliable [21]. By reducing the window size, e.g. from daily price and minimum daily price. There, storage charging/
24 hours to 1 hour, forecasts become more reliable [22], discharging rates are constraints in the optimization problem,
nevertheless storage management might be inefficient due to while storage dispatch depends on comparing the charging
the lack of information. So there is a trade-off between forecast price with local generation costs.
precision and EMS performance. Colson et al. in [26] developed a cost-efficient index for
In general, overall reliability is improved when: the impact optimal operation of the storage system. Again, this is imple-
of less variable inputs (e.g. loads) is greater than the more mented in an independent fashion and in addition offers a user
variable input (e.g. wind) [23]; and the forecasting window interface to decide if battery health is more important than the
is moved in order to always predict with the latest data [22]. microgrid’s revenue stream.
Of course, if two-day-ahead forecast error is close enough to
 3

X 
B. Reserve management F (~
pt , ~πt ) = a1 f1 + a2 f2 + . . . an fn an = 1 (2)
Reserve management is included in the Energy Managment Therefore, the new problem is to minimize F subject to the
System (EMS) as a constraint. For example, in [21] system constraints already mentioned –given a set of an coefficients.
reserve (including wind availability) is set to be 5% greater For each possible combination of coefficients, an optimal
than the demand. Another way is to split reserve into both, solution is found and then, a Pareto front point. This is not
operating and idle local generators, as proposed by [27]. On the only way to transform the original objective function (for
the other hand, when calculating the expected energy not a complete list refer again to [28]) but it seems to be the
supplied, in [23] it is included the reserve into the objective most common choice in almost all papers reviewed, even
function by establishing a reserve bid price. with decentralized solution methods. Parisio et al. provide
some support for this choice in [30]. Anyway, new lines of
V. O PTIMIZATION research on this topic may include exploring the behavior
The optimization problem implicit by an Energy Manage- of other objective functions transformations and exploiting
ment Systems (EMS) could be stated as [18]: and adapting the extra information of other methods, like the
ε−constrained method.
min {f1 (~
pt , ~πt ) , f2 (~
pt , ~πt ) , . . . fn (~
pt , ~πt )} (1)
Although the usual objectives are related to operational and
s.t. p~t ∈ P environmental costs [31]–[33], some authors add reliability-
∀t = 1, 2, . . . T based objectives [21], [23]. On the other hand, in reference
[34] it is concluded that less outage costs are incurred after
Where p~t are all the power injections by local generators
a short local generation interruption if continuous transitions
and the microgrid at time t, ~πt are all power prices at time t
to islanded mode are preferred over islanded operation. So a
–which are normally known, i.e. they are problem parameters,
simple operational policy can also be important for the overall
fk is the k cost function to be minimized, n are the number
objective function.
of cost functions (n ≥ 2), T are the number of time slots to
Fixed costs in the optimization problem can be included as
be considered and P is the set of all feasible p~t for any time
a penalty cost function that depends on each extra (unused)
slot. For a complete description of usual EMS multiobjective
energy unit. The idea is to make fixed cost recovery time
optimization constraints, including start-up and shut-down
longer if the energy is not used. This philosophy is imple-
costs, see [22]. For a formulation which includes transaction
mented in [17] as a penalty function based on the difference
costs see [18].
of the maximum generation available and actual generation.
Since all objectives have to be optimized at once, this is
The extra power is multiplied by a constant that represents
called a multiobjective or vector optimization. In contrast with
the fixed costs evenly distributed across the lifetime of the
standard optimization problems, these kind of problems do not
project. With a similar idea, [35] include a couple of penalty
have a single answer by themselves. Instead, a multiobjective
functions into the objective function of the EMS: the penalty
optimization solution gives all the possible optimal points
cost function depends on the excess in local wind generation
that meet the constraints. This set of points are called the
and the reserve cost function depends on the unavailability of
Pareto front, and the selection of one single point is left
local wind generation.
to the so called decision maker (for a detailed and rigorous
Optimization problem in [36] consists on minimizing energy
explanation of multiobjective optimization please refer to [28],
bill price in first place and then a voltage variability index
[29]). Two main approaches have been proposed to solve the
(based on steady state operation); the latter optimization flatten
EMS optimization problem:
voltage fluctuations between time slots. Salani et al. show that
Centralized calculation Optimization is done with all the in- cost minimization leaded with price signals impacts negatively
formation available and without a coordination on the voltage variability index, then imposing a physical limit
strategy between microgrid participants. A central to the price reactiveness of microgrid participants.
computation agent broadcast to each participant
the respective optimal set point.
B. Centralized solutions
Decentralized calculation Optimization is done with partial
information and a coordination strategy between Two kinds of solution methods to solve the EMS optimiza-
microgrid participants. A local computation pro- tion problem will be presented: methods based on heuristics
cess leaded by each participant determines the and non-heuristic methods.
respective optimal set point. 1) Heuristics: Two main heuristics are usually employed to
solve the optimization problem associated with the microgrid
Next, it is a discussion about the objective functions in EMS,
operation: Genetic Algorithms [25], [27], [35], [37], [38] and
followed by a description of some centralized and decentral-
Particle Swarm Optimization [20], [39]–[41]. In particular,
ized methods for EMS optimization problem solution.
reference [41] shows how Particle Swarm Optimization explo-
ration capabilities can be improved with Differential Evolution
A. Objective functions Algorithm.
A common practice to solve multiobjective optimization Two works from genetic algorithm implementations stands
problems is to replace the original objective function by a out: In [27] the genetic algorithm is modified by incorporating
convex combination of each objective: a Simulated Annealing (SA) heuristic to discard poor solutions.
 4

With SA the genetic algorithm gets a velocity boost (from one offers a way to analyze market operation, while the latter
nearly an hour to a couple of minutes) without major precision allow the simulation of smart grid concepts, like the «Virtual
lost (0,78% with respect the case without SA); In second Power Plant».
place, Xu et al. in [35] compare the performance of the Not only simulation efforts are made in the field, there are
Strength Pareto Evolutionary Algorithm 2 (SPEA2) and the some implementations of microgrids with MAS. A (mini)pilot
Non-dominated Sorting Genetic Algorithm (NSGAII), and with a Energy Management System based on MAS at Nanyang
found that SPEA2 is a faster (needs fewer generations to find Technological University is presented in [54]. Additional to
an optimal) but less precise method than NSGAII. the implementation results, the authors show an overview
2) Other methods: A control-theory like Energy Manage- of the MAS JADE interfaces. In [55] are briefly presented
ment System (EMS) is presented in [42] (these kind of al- the «GridAgents Framework» pilot ran in Australia and
gorithms are better known as Adaptive Model-based Control). the MORE-Microgrids project MAS pilot in Kythnos island
The method was tested with a long-term forecast window of 12 (Greece).
hours and a 20 minutes update rate, which yield to a reduction 2) Methods: As in [43], the objective functions in [56]
of 21% of the cost with respect to a standard EMS strategy. are separable, nonetheless instead of parallel computation and
Sathyanarayana et al. develop in [43] a Locational Marginal because separability is translated to the dual problem, in [56]
Price (LMP) to energy pricing in a power distribution system. the authors use a subgradient algorithm to do a distributed
This LMP is calculated for a specific Energy Management Sys- calculation of the optimum. Here, dual decomposition solution
tem (EMS) to minimize the peak power demand, the energy is possible because power from microgenerators is taken from
consumed, the total cost of energy and the total loss in the a time series, so it is a known value during the optimization
system. Because the objective functions can be expressed as a steps.
sum of simpler functions, the solution method is implemented However, as mentioned in [57], because dual decomposition
for parallel processing. message complexity is super-linear with the number of neigh-
Fathi et al. in [44] minimize the cost function and the bors and the maximum length path, researchers must be careful
peak to average ratio of the load profile considering imperfect when selecting the stabilizing step size of the messages,
information over the actual loads of the system. Although otherwise, the communication burden will be unacceptable.
the solution is suboptimal, Fathi et al. show that under load This is why in [58] the authors develop a decentralized
uncertainty the method outperforms game theory approaches. optimization method to solve the DC optimal power flow
problem that in addition does not require constant availability
C. Decentralized solutions of the communication network. Instead, it converges to an
optimum through an asynchronous sample of measures, using
In this section, first it will be presented researches which in-
the communication network only when there is an immediate
volve multiagent systems. Then, it will be presented researches
need.
that involve the application of decomposition methods to the
Another problem with decentralized optimization methods
energy management system problem.
is the rate of convergence. To solve this, Wei et al. in [59]
1) Multiagent systems: Microgrids are decentralized sys-
have proposed an asynchronous Alternating Direction Method
tems, so applying Multiagent Systems (MAS) tools for mi-
of Multipliers (ADMM) algorithm. Wei et al. probe that this
crogrid management seems natural –basic notions of agency,
method has a rate of convergence of O(1/k) –where √ k is the
MAS implementations and its relation with microgrids can be
number of iterations. Compare this with the O(1/ k) rate of
found in [45]–[48].
convergence of dual decomposition solution with subgradient
Colson et al. in [49] implement a MAS which manage a
methods and the O(1/k 2 ) rate of convergence of gradient
simulated microgrid with two main objectives, cost minimiza-
methods with global information [59]. Although Wei et al.
tion and performance maximization. Another implementation,
applies the method to a general network utility maximization
this time specifying the system models implemented and the
problem, the results can be implemented to solve a DC optimal
full MAS developing process is found in [50].
power flow problem.
Logenthiran et al. in [51] maximize microgrid profit by
Dual decomposition analysis also can be used to find an
minimizing the distance of the actual load profile and a
indicator function for the unit commitment problem, as shown
reference load profile. Authors in [51] left to future works the
in [60], thus reducing the complexity of the optimal energy
decision process for optimal reference load profile selection.
management in a microgrid.
Ren et al. in [52] present a (well documented) MAS tool for
Now, decentralized dispatch with losses is difficult because
microgrid dynamic management based on InterPSS (microgrid
system losses are a function of almost all power system ele-
environment), JADE (MAS handling) and JUNG (Network
ments. In reference [61], Binetti et al. solved this problem by
visualization). The optimum is achieved through a well defined
calculating the power mismatch in each node in a distributed
set of rules that determine agents’ interaction for each possible
fashion in parallel with a consensus algorithm which updates
scenario. The InterPSS integration offers not only a mean to
the power cost for each generator. Results show that the
compute power flow output but also a way to modeling and
method has a performance similar to the lambda method and
simulating connection and disconnection processes.
the reliability advantages of the decentralized computation.
A set of two MAS tools are employed in [53] to analyze
microgrid market. Namely, MASCEM and MASGriP: The first
 5

VI. D ISCUSSION [9] R. Lasseter, “Microgrids,” in Power Engineering Society Winter Meeting,
2002. IEEE, vol. 1, pp. 305–308 vol.1, 2002.
Decentralized solutions based in optimization problem de- [10] S. Chowdhury, S. Chowdhury, and P. Crossley, Microgrids and Active
composition often requires a good insight in convex pro- Distribution Networks. The Institution of Engineering and Technology,
gramming and modeling [62]–[64]. In this direction efforts 2009.
[11] A. Banerji, D. Sen, A. K. Bera, D. Ray, D. Paul, A. Bhakat, and S. K.
are beginning to extend the knowledge in power system Biswas, “Microgrid: A review,” in Global Humanitarian Technology
analysis [65]. Game-theoretic methods [66], [67], constraint- Conference: South Asia Satellite (GHTC-SAS), 2013 IEEE, pp. 27–35,
programming [68], adaptive stochastic control [69] (including 2013.
[12] S. Bhaskara and B. Chowdhury, “Microgrids a review of modeling,
pure approximate dynamical programming [70]) and com- control, protection, simulation and future potential,” in Power and
plementary systems analysis [71] are in an early stage of Energy Society General Meeting, 2012 IEEE, pp. 1–7, 2012.
development, so constitute an interesting research option. [13] I. Std., “Ieee guide for design, operation, and integration of distributed
resource island systems with electric power systems,” IEEE Std 1547.4-
From the review was observed a trend towards the concept 2011, pp. 1–54, July 2011.
of the Active Distribution Network [10], [72], [73]. There, the [14] H. Kirkham, D. Nightingale, and T. Koerner, “Energy management
Energy Management System becomes the Distributed Energy system design with dispersed storage and generation,” Power Apparatus
and Systems, IEEE Transactions on, vol. PAS-100, no. 7, pp. 3432–3441,
Resource Management System1 . This concept preserves the 1981.
philosophy behind an interconnected-mode microgrid with [15] H. Kanchev, D. Lu, B. Francois, and V. Lazarov, “Smart monitoring
several distributed energy resource and opens the perspective of a microgrid including gas turbines and a dispatched pv-based active
generator for energy management and emissions reduction,” in Innova-
towards a system wide analysis. tive Smart Grid Technologies Conference Europe (ISGT Europe), 2010
IEEE PES, pp. 1–8, 2010.
VII. C ONCLUSIONS [16] S. Conti and S. A. Rizzo, “Optimal control to minimize operating costs
and emissions of mv autonomous micro-grids with renewable energy
Energy Management Systems in a microgrid can lower sources,” in Clean Electrical Power, 2009 International Conference on,
the generation cost of microgenerators by allocating energy pp. 634–639, 2009.
[17] S. Chakraborty and M. Simoes, “Pv-microgrid operational cost mini-
resources in a power distribution system optimally. In this mization by neural forecasting and heuristic optimization,” in Industry
paper were presented an analysis of such system and some Applications Society Annual Meeting, 2008. IAS ’08. IEEE, pp. 1–8,
of the main strategies for its successful implementation. It can 2008.
[18] A. Chaouachi, R. Kamel, R. Andoulsi, and K. Nagasaka, “Multiobjective
be seen that future power distribution system problems are not intelligent energy management for a microgrid,” Industrial Electronics,
just a small version of power transmission problems. A mix IEEE Transactions on, vol. 60, no. 4, pp. 1688–1699, 2013.
of old and new tools have to be used in order to solve the new [19] C. Changsong, D. Shanxu, C. Tao, L. Bangyin, and Y. Jinjun, “Energy
trading model for optimal microgrid scheduling based on genetic al-
challenges that comes ahead. gorithm,” in Power Electronics and Motion Control Conference, 2009.
IPEMC ’09. IEEE 6th International, pp. 2136–2139, 2009.
R EFERENCES [20] T. Mahmoud, D. Habibi, and O. Bass, “Fuzzy logic for smart utilisation
of storage devices in a typical microgrid,” in Renewable Energy Research
[1] M. Hoogwijk and W. Graus, “Global potential of renewable energy and Applications (ICRERA), 2012 International Conference on, pp. 1–6,
sources: A literature assesment,” tech. rep., ECOFYS: Sustainable en- 2012.
ergy for everyone, 2008. [21] M. Meiqin, J. Meihong, D. Wei, and L. Chang, “Multi-objective eco-
[2] D. Schnitzer, D. S. Lounsbury, J. P. Carvallo, R. Deshmukh, J. Apt, and nomic dispatch model for a microgrid considering reliability,” in Power
D. M. Kammen, “Microgrids for rural electrification: A critical review Electronics for Distributed Generation Systems (PEDG), 2010 2nd IEEE
of best practices based on seven case studies,” tech. rep., United Na- International Symposium on, pp. 993–998, 2010.
tions Foundation, University of California (Berkeley), Carnegie Mellon [22] R. Palma-Behnke, C. Benavides, F. Lanas, B. Severino, L. Reyes,
University, 2014. J. Llanos, and D. Saez, “A microgrid energy management system based
[3] M. Barnes, J. Kondoh, H. Asano, J. Oyarzabal, G. Ventakaramanan, on the rolling horizon strategy,” Smart Grid, IEEE Transactions on,
R. Lasseter, N. Hatziargyriou, and T. Green, “Real-world microgrids-an vol. 4, no. 2, pp. 996–1006, 2013.
overview,” in System of Systems Engineering, 2007. SoSE ’07. IEEE [23] M. Q. Wang and H. Gooi, “Effect of uncertainty on optimization of
International Conference on, pp. 1–8, April 2007. microgrids,” in IPEC, 2010 Conference Proceedings, pp. 711–716, 2010.
[4] M. Agrawal and A. Mittal, “Microgrid technological activities across [24] R. Karki, S. Thapa, and R. Billinton, “A simplified risk-based method
the globe: A review,” International Journal of research and reviews in for short-term wind power commitment,” Sustainable Energy, IEEE
applied sciences, vol. 7, pp. 147–152, 2011. Transactions on, vol. 3, no. 3, pp. 498–505, 2012.
[5] W. Vergara, P. Isbell, A. Rios, J. R. Gómez, and L. Alves, “Beneficios [25] C. Chen, S. Duan, T. Cai, B. Liu, and G. Hu, “Smart energy management
para la sociedad de la adopción de fuentes renovables de energía system for optimal microgrid economic operation,” Renewable Power
en américa latina y el caribe,” tech. rep., Banco Interamericano de Generation, IET, vol. 5, no. 3, pp. 258–267, 2011.
Desarrollo, División de Cambio Climático y Sostenibilidad, División [26] C. Colson, M. Nehrir, R. Sharma, and B. Asghari, “Improving sustain-
de Energía, Nota Técnica No. IDB-TN-623., Abril 2014. ability of hybrid energy systems part i: Incorporating battery round-trip
[6] G. Morris, C. Abbey, S. Wong, and G. Joos, “Evaluation of the costs efficiency and operational cost factors,” 2013.
and benefits of microgrids with consideration of services beyond energy [27] H. Z. Liang and H. Gooi, “Unit commitment in microgrids by improved
supply,” in Power and Energy Society General Meeting, 2012 IEEE, genetic algorithm,” in IPEC, 2010 Conference Proceedings, pp. 842–
pp. 1–9, 2012. 847, 2010.
[7] D. Gazzoni, I. Azurdia, G. Blanco, C. Estrada, and I. d. C. Macedo, [28] K. Miettinen, Nonlinear Multiobjective Optimization. Kluwer Academic
“Sustainable energy in latin america and the caribbean: Potential for the Publishers, Operations Research and Management Science, 1999. Re-
future,” tech. rep., International Council for Science, Regional Office for visión más extensa y profunda de todos los métodos.
Latin America and the Caribbean, 2010. [29] Y. Sawaragi, H. Nakayama, and T. Tanino, Theory of Multiobjective
[8] F. Katiraei, R. Iravani, N. Hatziargyriou, and A. Dimeas, “Microgrids Optimization. Academic Press, Inc., 1985. Muy elevado el nivel, pero
management,” Power and Energy Magazine, IEEE, vol. 6, pp. 54–65, explica de manera razonable la custión de las estructuras de dominación.
May 2008. [30] A. Parisio and L. Glielmo, “Multi-objective optimization for envi-
ronmental/economic microgrid scheduling,” in Cyber Technology in
1 Term used in the last call for papers of the IEEE TRANSACTIONS ON Automation, Control, and Intelligent Systems (CYBER), 2012 IEEE
POWER DELIVERY Special Issue on «Protection and Real-Time Monitoring International Conference on, pp. 17–22, 2012.
of Transmission and Distribution Systems with High Penetration of Distributed [31] F. Farahmand, T. Khandelwal, J. J. Dai, and F. Shokooh, “An enterprise
Generati on and Microgrids». approach to the interactive objectives and constraints of smart grids,” in
 6

Innovative Smart Grid Technologies - Middle East (ISGT Middle East), [53] P. Oliveira, T. Pinto, I. Praca, Z. Vale, and H. Morais, “Intelligent
2011 IEEE PES Conference on, pp. 1–6, 2011. micro grid management using a multi-agent approach,” in PowerTech
[32] C. Colson and M. Nehrir, “A review of challenges to real-time power (POWERTECH), 2013 IEEE Grenoble, pp. 1–6, 2013.
management of microgrids,” in Power Energy Society General Meeting, [54] F. Eddy and H. B. Gooi, “Multi-agent system for optimization of
2009. PES ’09. IEEE, pp. 1–8, 2009. microgrids,” in Power Electronics and ECCE Asia (ICPE ECCE), 2011
[33] E. Sortomme and M. El-Sharkawi, “Optimal power flow for a system IEEE 8th International Conference on, pp. 2374–2381, 2011.
of microgrids with controllable loads and battery storage,” in Power [55] R. Roche, B. Blunier, A. Miraoui, V. Hilaire, and A. Koukam, “Multi-
Systems Conference and Exposition, 2009. PSCE ’09. IEEE/PES, pp. 1– agent systems for grid energy management: A short review,” in IECON
5, 2009. 2010 - 36th Annual Conference on IEEE Industrial Electronics Society,
[34] D. Pudjianto, E. Zafiropoulos, and L. Daoutis, “Methodology for quan- pp. 3341 –3346, nov. 2010.
tifying economic and environmental benefits of microgrids,” tech. rep., [56] Y. Zhang, N. Gatsis, and G. Giannakis, “Robust energy management
Large scale integration of microgrids to low voltage grids, deliverable for microgrids with high-penetration renewables,” Sustainable Energy,
DG4, 2005. IEEE Transactions on, vol. 4, no. 4, pp. 944–953, 2013.
[35] J. Xu, S. Tan, and S. Panda, “Optimization of economic load dispatch [57] P. Wan and M. Lemmon, “Distributed network utility maximization
for a microgrid using evolutionary computation,” in IECON 2011 - 37th using event-triggered augmented lagrangian methods,” in American
Annual Conference on IEEE Industrial Electronics Society, pp. 3192– Control Conference, 2009. ACC ’09., pp. 3298–3303, 2009.
3197, 2011. [58] P. Wan and M. Lemmon, “Optimal power flow in microgrids using event-
[36] M. Salani, A. Giusti, G. Di Caro, A.-E. Rizzoli, and L. Gambardella, triggered optimization,” in American Control Conference (ACC), 2010,
“Lexicographic multi-objective optimization for the unit commitment pp. 2521–2526, 2010.
problem and economic dispatch in a microgrid,” in Innovative Smart [59] E. Wei and A. Ozdaglar, “On the o(1/k) convergence of asynchronous
Grid Technologies (ISGT Europe), 2011 2nd IEEE PES International distributed alternating direction method of multipliers,” Mathematical
Conference and Exhibition on, pp. 1–8, 2011. Programming, p. preprint, 2013.
[37] B. Zhao, X. Zhang, J. Chen, C. Wang, and L. Guo, “Operation opti- [60] B. Zhao, Y. Shi, X. Dong, W. Luan, and J. Bornemann, “Short-term
mization of standalone microgrids considering lifetime characteristics operation scheduling in renewable-powered microgrids: A duality-based
of battery energy storage system,” 2013. approach,” Sustainable Energy, IEEE Transactions on, vol. 5, pp. 209–
[38] H. Vahedi, R. Noroozian, and S. Hosseini, “Optimal management of 217, Jan 2014.
microgrid using differential evolution approach,” in Energy Market [61] G. Binetti, A. Davoudi, F. Lewis, D. Naso, and B. Turchiano, “Dis-
(EEM), 2010 7th International Conference on the European, pp. 1–6, tributed consensus-based economic dispatch with transmission losses,”
2010. Power Systems, IEEE Transactions on, vol. 29, pp. 1711–1720, July
[39] A. Basu, S. Chowdhury, and S. P. Chowdhury, “Operational management 2014.
of chp-based microgrid,” in Power System Technology (POWERCON), [62] S. Low, D. Gayme, and U. Topcu, “Convexifying optimal power
2010 International Conference on, pp. 1–5, 2010. flow: Recent advances in opf solution methods,” in American Control
[40] C. xia Dou, X. bei Jia, Z. qian Bo, F. Zhao, and D. le Liu, “Optimal man- Conference (ACC), 2013, pp. 5245–5245, June 2013.
agement of microgrid based on a modified particle swarm optimization [63] B. Lesieutre, D. Molzahn, A. Borden, and C. DeMarco, “Examining
algorithm,” in Power and Energy Engineering Conference (APPEEC), the limits of the application of semidefinite programming to power flow
2011 Asia-Pacific, pp. 1–8, 2011. problems,” in Communication, Control, and Computing (Allerton), 2011
[41] N. Bazmohammadi, A. Karimpour, and S. Bazmohammadi, “Optimal 49th Annual Allerton Conference on, pp. 1492–1499, Sept 2011.
operation management of a microgrid based on mopso and differential [64] E. Dall’Anese, H. Zhu, and G. Giannakis, “Distributed optimal power
evolution algorithms,” in Smart Grids (ICSG), 2012 2nd Iranian Con- flow for smart microgrids,” Smart Grid, IEEE Transactions on, vol. 4,
ference on, pp. 1–6, 2012. pp. 1464–1475, Sept 2013.
[42] A. Hooshmand, B. Asghari, and R. Sharma, “A novel cost-aware multi- [65] J. Lavaei, D. Tse, and B. Zhang, “Geometry of power flows and opti-
objective energy management method for microgrids,” in Innovative mization in distribution networks,” Power Systems, IEEE Transactions
Smart Grid Technologies (ISGT), 2013 IEEE PES, pp. 1–6, 2013. on, vol. 29, pp. 572–583, March 2014.
[43] B. Sathyanarayana and G. Heydt, “Sensitivity-based pricing and opti- [66] W. Saad, Z. Han, H. Poor, and T. Basar, “Game-theoretic methods for
mal storage utilization in distribution systems,” Power Delivery, IEEE the smart grid: An overview of microgrid systems, demand-side man-
Transactions on, vol. 28, no. 2, pp. 1073–1082, 2013. agement, and smart grid communications,” Signal Processing Magazine,
[44] M. Fathi and H. Bevrani, “Adaptive energy consumption scheduling for IEEE, vol. 29, pp. 86–105, Sept 2012.
connected microgrids under demand uncertainty,” Power Delivery, IEEE [67] H. Soliman and A. Leon-Garcia, “Game-theoretic demand-side manage-
Transactions on, vol. 28, no. 3, pp. 1576–1583, 2013. ment with storage devices for the future smart grid,” Smart Grid, IEEE
[45] R. Bordini, J. Hubner, and M. Wooldridge, Programming Multi-Agent Transactions on, vol. 5, pp. 1475–1485, May 2014.
Systems in AgentSpeak using Jason. Wiley Series in Agent Technology, [68] M. Dolan, E. Davidson, G. Ault, K. Bell, and S. McArthur, “Distribution
Wiley, 2007. power flow management utilizing an online constraint programming
[46] S. McArthur, E. Davidson, V. Catterson, A. Dimeas, N. Hatziargyriou, method,” Smart Grid, IEEE Transactions on, vol. 4, pp. 798–805, June
F. Ponci, and T. Funabashi, “Multi-agent systems for power engineering 2013.
applications; part i: Concepts, approaches, and technical challenges,” [69] R. Anderson, A. Boulanger, W. Powell, and W. Scott, “Adaptive
Power Systems, IEEE Transactions on, vol. 22, pp. 1743 –1752, nov. stochastic control for the smart grid,” Proceedings of the IEEE, vol. 99,
2007. pp. 1098–1115, June 2011.
[47] S. McArthur, E. Davidson, V. Catterson, A. Dimeas, N. Hatziargyriou, [70] M. Strrelec and J. Berka, “Microgrid energy management based on
F. Ponci, and T. Funabashi, “Multi-agent systems for power engineering approximate dynamic programming,” in Innovative Smart Grid Tech-
applications; part ii: Technologies, standards, and tools for building nologies Europe (ISGT EUROPE), 2013 4th IEEE/PES, pp. 1–5, Oct
multi-agent systems,” Power Systems, IEEE Transactions on, vol. 22, 2013.
pp. 1753 –1759, nov. 2007. [71] S. Gabriel, A. Conejo, J. Fuller, B. Hobbs, and C. Ruiz, Complementarity
[48] M. Wooldridge, An Introduction to MultiAgent Systems. Wiley, 2009. Modeling in Energy Markets. Springer, 2012.
[49] C. Colson, M. Nehrir, R. Sharma, and B. Asghari, “Improving sustain- [72] S. Gill, I. Kockar, and G. Ault, “Dynamic optimal power flow for active
ability of hybrid energy systems part ii: Managing multiple objectives distribution networks,” Power Systems, IEEE Transactions on, vol. 29,
with a multiagent system,” 2013. pp. 121–131, Jan 2014.
[50] C. Colson and M. Nehrir, “Comprehensive real-time microgrid power [73] Z. Hu and F. Li, “Cost-benefit analyses of active distribution network
management and control with distributed agents,” Smart Grid, IEEE management, part i: Annual benefit analysis,” Smart Grid, IEEE Trans-
Transactions on, vol. 4, no. 1, pp. 617–627, 2013. actions on, vol. 3, no. 3, pp. 1067–1074, 2012.
[51] T. Logenthiran, D. Srinivasan, A. Khambadkone, and H. N. Aung,
“Multiagent system for real-time operation of a microgrid in real-time
digital simulator,” Smart Grid, IEEE Transactions on, vol. 3, no. 2,
pp. 925–933, 2012.
[52] F. Ren, M. Zhang, and D. Sutanto, “A multi-agent solution to distribu-
tion system management by considering distributed generators,” Power
Systems, IEEE Transactions on, vol. 28, no. 2, pp. 1442–1451, 2013.

View publication stats