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

net/publication/268238044

SGsim: A simulation framework for smart grid applications

Conference Paper · May 2014

DOI: 10.1109/ENERGYCON.2014.6850507

CITATIONS READS
30 866

3 authors:

Abdalkarim Awad Peter Bazan

Birzeit University Friedrich-Alexander-University of Erlangen-Nürnberg
46 PUBLICATIONS 612 CITATIONS 51 PUBLICATIONS 441 CITATIONS

SEE PROFILE SEE PROFILE

Reinhard German
Friedrich-Alexander-University of Erlangen-Nürnberg
289 PUBLICATIONS 6,499 CITATIONS

SEE PROFILE

All content following this page was uploaded by Abdalkarim Awad on 12 October 2015.

The user has requested enhancement of the downloaded file.

 SGsim: a Simulation Framework for Smart Grid
Applications
Abdalkarim Awad, Peter Bazan and Reinhard German
Computer Networks and Communication Systems
University of Erlangen, Germany
{abdalkarim.awad,peter.bazan,german}@cs.fau.de

Abstract—Simulation tools are an essential component of

any emerging technology. In this paper we present SGsim, Control
Market
a framework to simulate different applications in the context
of smart grid. The framework supports real time simulation. Data
Therefore it is possible to evaluate time-critical applications such Management
as real time monitoring and control. The framework combines
two main simulators (1) OMNeT++, a discrete event simulator
that is used mainly to simulate data communication systems and Smart
art Gri
Grid
rid
ri
Power
(2) OpenDSS, a tool to calculate the power flow in power grids. Grid
Moreover, the framework supports smart grid related standards
such as IEEE C37.118. This way it is possible to integrate
Communication
standard smart grid tools such as openPDC. Furthermore, an
optimization toolbox is integrated in the simulator in addition to
the capability to communicate with other tools such as MATLAB
and R. We performed a set of case studies to show the capabilities Optimization
of the simulator. Security and Privacy
Index Terms—Co-simulation, Smart grid, OMNeT++,
OpenDSS, openPDC, Communication standards.
Fig. 1. The multi-disciplinary field of smart grid

I. I NTRODUCTION
Smart grid is a broad and multi-disciplinary field as shown • application that need to deal with standards
in Figure 1. Therefore, dealing with new approaches and There are several simulation tools for the analysis of smart
algorithm in the context of smart grid may require dealing grids with renewable energy sources. The Integrated Simu-
with diversity of fields such as power analysis, communication, lation Environment Language (INSEL) [1] is a discrete time
control, optimization, data management, etc. To realize the step simulation tool for renewable energy systems. It is written
smart grid, many innovative architectures, control mechanisms in Fortran and C++, whereas the graphical user interface
and network technologies are required, for instance volt-var is implemented in Java. It has many predefined models of
optimization, distributed generation and storage, and quasi-real power system components integrated, which can be connected
time monitoring and control of the power grid. To evaluate to complex power grids. INSEL can communicate over the
the impact and effectiveness of the proposed approaches in internet with real world systems but has no built in communi-
addressing the aforementioned issues, simulation plays a key cation protocols for simulating the communication between
role because a real power grid can not be used for testing the simulated models. Furthermore, it has no optimization
and validation. Obviously, simulation in both the areas of integrated. A hybrid simulation framework for smart grids with
communication networks as well as power systems has been renewable energy systems is i7-AnyEnergy [2], [3]. It is based
widely adopted. However, the coupling of those two worlds in on AnyLogic [4] and uses the system dynamic paradigm for
the frame of smart grids calls for tools that are able to address energy flows and the discrete event simulation paradigm for
both worlds. control decisions. Components can communicate by sending
The main contribution of this work is a framework that messages over ports, but no standard communication protocols
supports different types of smart grid applications such as are integrated. EnergyPLAN [5] is written in Delphi Pascal and
• applications that require only power flow for the optimization of energy systems with renewable energy
• applications that require detailed voltage and phase mea- sources with an hourly resolution. Beneath electricity grids, it
surements includes gas grids as well as heating and cooling grids. Instead
• application that require real time data communication of integrated stochastic models of the energy production or
and hence detailed simulation models of communication demand, the hourly distribution of the total annual production
components are important or demand has to be specified. The focus is on optimization
• optimization based applications of the operation of the specified energy system. It cannot
 OpenPDC
SuperPDC
TCP/IP
Opt.
PDC PMU OpenDSS
IEEE C Scripts OpenDSS DLL Interface
OMNeT++ (Power
C37.118 O
Simulator) PDC PDC
M
Inter
-face Script
INET Framework (Circuit
Description) PMU PMU PMU PMU PMU

Input/ Output

(a) (b) (c)

Fig. 2. Components of the SGsim Framework (a) openDSS power simulator (b), and Phasor Data Concentrator (PDC) (c).

communicate with real world systems and does not consider power Distribution System Simulator (DSS) for supporting
communication protocols or the electrical system. Renewable distributed resource integration and grid modernization efforts.
energy systems with electricity and heat generation/demand This simulator is used to simulate electric power networks.
together with economic constraints can be analyzed with It contains predefined models for different power elements
TIMES (The Integrated MARKAL-EFOM System) [6]. It is such as transformers, transmission lines and generators. To
programmed in GAMS [7], the General Algebraic Modeling simulate the communication part of the smart grid, the open
System for mathematical programming and optimization. With source discrete event simulator (OMNeT++) [9] is used. Inside
a linear-programming approach, the model of the energy OMNeT++ we have built some smart grid components such
system is solved with respect to least cost. The time resolution as Phasor Measurement Units (PMU) and a Phasor Data
of the generation/demand time series input data ranges from Concentrator (PDC) interface that can communicate with a real
day and night to a year. TIMES doesn’t consider the electrical PDC such as openPDC [14]. The openPDC administered by
system or communication protocols. A communication and the Grid Protection Alliance (GPA) is a complete Phasor Data
power distribution network co-simulation tool for smart grids Concentrator software system designed to process streaming
is presented in [8]. The realization with OMNeT++ [9] and time-series data in real-time. The openPDC implements a
OpenDSS [10] is described and a case study is given. The number of standard phasor protocols, which can be used
co-simulator is a hybrid simulation framework because of to receive data from devices. The supported protocols are
the two involved simulators. OMNeT++ allows the use of IEEE C37.118 , IEEE 1344, BPA PDCstream, FNET, SEL
standard communication protocol models. An optimization or Fast Message, and Macrodyne. The IEEE 1344 standard for
a communication with real world systems is not implemented. synchrophasors was completed in 1995 and reaffirmed in 2001.
Another co-simulation tool is GridLAB-D [11]. Its agent-based In 2005, it was replaced by IEEE C37.118-2005 [15], which
simulation engine is implemented in C++. The communication was a complete revision and dealt with issues concerning use
network is modelled as a black box with simplified network of PMUs in electric power systems. The INET framework
properties like bandwidth, latency, congestion, or buffer size. [16] implements several detailed data communication protocol
There are no standard communication protocols available. In such as TCP/IP and devices such as routers and hosts. Another
older versions of the tool, no optimization was available but component of SGsim is the optimization unit. This unit makes
newer versions make use of the optimization tool MatPOWER it possible to explore optimized approaches inside the smart
[12]. GridLAB-D itself is used by the tool GridSpice [13] grid such as economical dispatch and profit maximization.
in conjunction with a transmission and economic dispatch
package for optimization based on MatPOWER. It is written in A. OMNeT++
Java and can distribute the computational effort for large sim- OMNeT++ is an open source discrete event simulator and
ulations over virtual machines using Amazon Web Services. is free for academic use. In fact there is a commercial
As with GridLAB-D, it does not come with built-in standard version of OMNeT++ called OMNEST. OMNeT++ It is
communication protocols. an object-oriented component-based discrete event network
The rest of the paper is organized as follows. In Section II, simulation framework which is gaining wide acceptance in
we present our framework, in Section III, we present several the scientific community for building and simulating commu-
case studies to show the basic capabilities of the the frame- nication systems. It provides the necessary tools to simulate
work. Finally, Section IV concludes the paper. different types of systems, but itself does not provide any
specific component for computer networks or any other area.
II. SG SIM Nevertheless, there are some extensions, such as the INET
The SGsim consists mainly of two main simulators and a framework [16], MiXiM framework [17] and the Mobility
standard software suit as shown in Figure 2(a). It uses the open framework [18], which have been released under the GNU
source power simulator OpenDSS [10], which is an electric General Public License (GPL) to support specific application
 (a) (b) (c)

Fig. 3. Snapshot from OMNeT++ of the simulation playground (with spapshot from google maps) (a) the PDC unit (b), and the PMU unit (c).

areas. A typical simulation model consists of modules that found in the program were originally intended to support
communicate through message passing. The simple modules distributed generation analysis needs. Other features support
are written in C++ and can be grouped into compound modules energy efficiency analysis of power delivery, smart grid ap-
and so forth. OMNeT++ models are often referred to as plications, and harmonics analysis. OpenDSS can be used for
networks and are described in OMNeT++’s NED (Network different purposes such as distribution planning and analysis,
Description) language. The initial value of the parameters can solar and wind plants simulations and storage modeling. The
be assigned either in the NED files or in the configuration file program has several built in solution modes such as snap
(omnetpp.ini). shot power flow and harmonics. Analyzing power systems
The INET framework contains detailed implementations using OpenDSS requires writing a script that describes the
for several protocols in the different network layers. For power system. OpenDSS supports a Component Object Model
example, it has implementations for TCP, UDP, IPv4 and (COM) interface which makes it possible to design and execute
IPv6. It provides implementations for link-layer models like custom solution modes and features from an external program
PPP, Ethernet and 802.11. Moreover it has implementations and perform the functions of the simulator, including definition
for different types of mobility. These implementations make it of the model data [19].
possible for users to implement their own protocols with little
effort on issues related to other layers. For example, to test C. OpenPDC
a routing protocol for Ad-hoc networks, the user can rely on A Phasor Data Concentrator (PDC) forms a node in a
the existing implementations for MAC and physical layers in system where phasor data from a number of PMUs or PDCs is
the simulation and evaluation of the protocol. Figure 3 shows correlated and fed out as a single stream to other applications.
the simulation environment in OMNeT++. Figure 3(a) shows The PDC correlates phasor data by time-tag to create a system
a snap shot from OMNeT++ of the simulation playground, wide measurement set. The PDC provides additional functions
Figure 3(b) shows the PDC unit, which uses UDP in the as well. It performs various quality checks on the phasor data
transport layer, and Figure 3(c) shows the PMU unit, which and inserts appropriate flags into the correlated data stream. It
also uses UDP in the transport layer. checks disturbance flags and records files of data for analysis.
It also monitors the overall measurement system and provides
B. OpenDSS a display and record of the performance. It can provide a
The Open Distribution System Simulator (OpenDSS, or number of specialized outputs, such as a direct interface to
simply, DSS) is a comprehensive electrical system simulation a Supervisory Control and Data Acquisition (SCADA) or
tool for electric utility distribution systems. OpenDSS refers to an Energy Management System (EMS). Figure 2(c) shows a
the open source implementation of the DSS. It is implemented common architecture of phasor network, where a number of
as both a standalone executable program and an in process PMUs send data to a common PDC and several PDCs, which
COM server DLL designed to be driven from a variety of belong to different utilities, are connected to a central PDC
existing software platforms. The executable version has a basic (SuperPDC).
text based user interface for the solution engine to assist The open source phasor data concentrator (openPDC) is
users in developing scripts and viewing solutions. The program a system, which is used to manage, process and respond to
supports nearly all rms steady state (i.e., frequency domain) dynamic changes in fast moving streaming phasor data. More
analyses commonly performed for utility distribution system specifically, the openPDC can process any kind of data that
planning and analysis. In addition, it supports many new types can be described as time-stamped measured values. These
of analyses that are designed to meet future needs, many measured values are simply numeric quantities that have been
of which are being dictated by the deregulation of utilities acquired at a source device and are typically called points,
worldwide and the advent of smart grid. Many of the features signals, events, time-series values or measurements. Examples
 G1 G2 L2

500
L2
B1
B2 L3

400
L5

Power (MW)

300
200
B5
B3

100
B4

0
G3 L5 L3 Wind
PV
0 5 10 15 20

Time (hours)

(a) (b) (c)

Fig. 4. Simple power system (a), snapshot of the system in OMNeT++ (b) and load on the different buses (c).
500

500

500
PV G1 P13
Wind G2 P23
400

400

400
G3 P53
Power (MW)

Power (MW)

Power (MW)
300

300

300
200

200

200
100

100

100
0

0
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

Time (hours) Time (hours) Time (hours)

(a) (b) (c)

Fig. 5. Produced power by PV and Wind (a), output from the generators without any limitations on the capacity of the transmission lines (a), power flow
from the different buses towards bus B3(b)
500

500

10000
G1 P13 w/o Limits
G2 P23 With Limits
400

400

G3 P53

Costs (Euro)
Power (MW)

Power (MW)
300

300

6000
200

200
100

100

2000
0

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

Time (hours) Time (hours) Time (hours)

(a) (b) (c)

Fig. 6. Output from the generators when the capacity of the transmission line between bus B1 and B3 is 200 MW (a), Power flow from the different buses
towards bus B3 when the maximum capacity of the transmission line between bus B1 to B3 is 200 MW (b) and comparison of the costs with and without
limitations on the capacity of the transmission line between bus B1 and B3 (c).

of measurements include temperature, voltage, vibration, lo- a method to quantify the measurement, tests to be sure the
cation, luminosity and, of course, phasors. When a value gets measurement conforms to the definition, and error limits
measured, an exact time stamp is taken, typically using a GPS- for the tests. It also defines a data communication protocol,
clock for accuracy. The value, along with its timestamp, is then including message formats for communicating this data in
streamed to the openPDC where it can be time-aligned with a real-time system [15]. Synchrophasor measurements shall
other incoming measurements so that an action can then be be synchronized to UTC time with sufficient accuracy. The
taken on a complete slice of data that was all measured at the standard proposes that the system must be capable of receiving
exact same moment in time. time from a highly reliable source, such as the Global Position-
ing System (GPS), which can provide sufficient time accuracy.
D. IEEE 37.118 All message frames start with a 2-byte SYNC word (0xAA
Currently, we implemented only the IEEE c37.118 standard and 8 bits which define the type of the frame) followed by a 2-
inside the framework, nevertheless, it is possible to add any byte FRAMESIZE word, a 2-byte IDCODE, and a time stamp
required standard. The standard defines synchronized phasor consisting of a 4-byte second of century. All frames terminate
measurements used in power system applications. It provides
with a check word (CHK), which is a CRC-CCITT. This CRC- B3 without considering limitations on transmission lines. As
CCITT uses the generating polynomial X 16 + X 12 + X 5 + 1 can be seen from the figure, most of the power comes from
with an initial value of 1 (hex FFFF). The SYNC word is generator one because it has the minimum costs. Moreover,
transmitted first and the check word last. Two- and 4-byte during the period of the day when the demand L3 is high,
words are transmitted most significant byte first (network or the power flow from bus B1 to B3 is more than 200 MW.
big endian order). Values of phasor and frequency can be send Then we have imposed a limitation on the transmission line
either using 16 bit integer or 32 float format. 13 between bus B1 and B3 such that the power is not allowed
Implementing this standard makes it possible to integrate to exceed 200M W . The output of the three generators as well
already existing software that also understands this standard. as the power flow towards bus B3 are shown in Figure 6(a) and
For example, using this standards, it is possible to send data Figure 6(b) respectively. As can be seen from the figures, the
from a phasor measurement unit to openPDC. power flow from bus B1 to B3 does not exceed the 200M W ,
nevertheless the power is coming now from more expensive
E. Optimization Tools generators. Figure 6(c) compares the costs of both cases.
One of the main characteristics of the smart grid is to
make the operation of the different entities of the smart grid
more efficient. To achieve this goal, optimization can play an G1(P 1) = 200 + 10P 1 + 0.00134P 12 100 < P 1 < 500
important role. In the SGsim framework, we integrated the (1)
open source optimization tools NLopt [20] and lpSolve [21].
NLopt is a free/open-source library for nonlinear optimization,
providing a common interface for a number of different G2(P 2) = 617 + 11P 2 + 0.00149P 22 0 < P 2 < 200 (2)
free optimization routines available online as well as original
implementations of various other algorithms. LpSolve is freely
available (under LGPL 2) software for solving linear, integer G3(P 3) = 400 + 12P 3 + 0.0020P 32 0 < P 3 < 500 (3)
and mixed integer programs. It is possible also to use the
optimization toolbox inside other tools such as MATLAB [22] B. Demand Response and Electric Vehicle
and R [23].
In the second case study we take the following simple
III. C ASE S TUDIES scenario, a house is a participant of the smart grid and it owns
a PV and an electric vehicle (EV). We are interested in an
In this section we use our simulator to test different smart
answer to the following question: what is the best strategy to
grid applications. In the first case study we test the economical
follow in order to maximize the profit and to have the car
dispatch together with optimal power flow. Then we test
charged when it is needed.
demand response in the second case study, in which we assume
To identify the optimal daily operational strategy to be
that the charging of an electric vehicle is an elastic load.
followed by a household, the day-ahead electricity price, a
Finally we study the wide area monitoring.
prediction of daily solar power and a prediction of required
A. Economical dispatch and optimal power flow energy demand are needed. We assume that a prediction of
supply, demand as well as the day-ahead price are avail-
The main goal of this case study is to show the ability of
able. The household maximizes the corresponding revenue
SGsim to study power systems only from power flow point
by exchanging power with the grid and charging at a cheap
of view and without considering voltages and currents. In the
price. Furthermore, households are charged for their power
first case study we have used the simple power system shown
consumption at market price and sell their power at feed-in
in figure 4(a) to study the economical dispatch and optimal
tariff.
power flow. All transmission lines have the same impedance
(Z = 0.1jΩ). Figure 4(b) shows the system in OMNeT++. The main idea is to use the day-ahead price to find the
The system consists of three generators which have different optimal strategy to be followed to maximize the profit of a
cost functions and limitations as shown in Eq 1, 2, and 3. household. When a household receives the day-ahead price it
We can impose limitations on the capacity of the transmission solves the optimization problem below for the whole day (i.e.
lines. We consider three load curves (L2, L3, L5) as shown T = 24 hours). It finds the optimal time to charge the EV-
in figure 4(c). Additionally, there are two renewable energy battery and what should be done with the power produced by
sources (PV and Wind) with production curves as shown in the solar system. Because the day-ahead price is available for
Figure 5(a) . In each step we match the supply with the each hour during the day, we selected a time step τ = 1 hour.
demand with minimum cost in addition to the limitations on The optimization problem can be defined as follows: The
the transmission lines. objective function is
At the beginning we looked at the economical dispatch
without considering any limitations on the capacity of the 23

transmission lines. Figure 5(b) and Figure 5(c) show the output max (Cf (t)τ Ps (t) − C(t)τ Pb (t)) (4)
of each generator as well as the power flow towards bus t=0
 1.5

0.5

10
Load Buy

Power (kW), SOC (kWh)

PV Sell

0.4

8
Charge

Price (Euro/kWh)
1.0
Battery SOC

Power (kW)

0.3

6
0.2

4
0.5

0.1

2
0.0

0.0

0
0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

Time (hours) Time (hours) Time (hours)

(a) (b) (c)

Fig. 7. House load and supply forecast (a), electricity price (b), and the amount of power that should be bought, sold, charged and the state of charge of
the EV-Battery (c).

The above maximization problem is subjected to system TABLE I

constraints (power balance), which can be written as, PARAMETERS USED IN THE CASE STUDY

Pv (t) − Pl (t) − Pc (t) + Pb (t) − Ps (t) = 0 (5) Charging efficiency of the battery (ηc ) 90%
Feed-in tariff (Cf ) 17 cent
where Cf is feed-in tariff, Pv is the power generated by Max charge power of battery (Pcmax ) 2kW
Storage initial energy (Estart ) 0 kWh
photovoltaic at the house, P c is the power charged in the Storage capacity (Emax ) 10 kWh
battery, Pl (t) represents the load at the house, C represents Required energy in the battery (Ereq ) 9 kWh
the market price, P b denotes the amount of power bought from Time to use the EV (R) 8 AM
the network, and P s denotes the amount of power sold to
the network. The hourly energy balance in the battery can be
modeled as: the PMU sends the magnitude and phasor at a rate of 25
E(t + 1) − E(t) − ηc τ Pc (t) = 0 (6) Frames/sec to the PMU connection Tester. The packets are
sent through communication channels that are simulated using
Pcmax≥ Pc (t) (7)
the INET-Framework. In this case study, the packets are send
E(0) = Estart (8) through an LTE communication channel as shown in Figure 3.
E(R) = Ereq (9) A screen shot of PMU connection tester is shown in Figure
Emax ≥ E(t) (10) 8. The PMU in the simulator sends a configuration frame
using IEEE c37.118 standard, where several parameters are
E(t) is the energy storage level in the battery at each hour. set such as nominal frequency (50 Hz), name and ID of the
Estart and Ereq are the initial and final energy storage levels PMU (PMU1 and V1), number of phasors (6), data rate (25
(required level). η c is the charge efficiency of the battery. R frames/sec), etc. Then the PMU starts sending the data frames,
is the time when the EV should be charged to the required which include the data elements such as the magnitude and
value. phase of the different signals, and time stamp. As it can be
Table I shows the parameters that have been used during this seen on the figure, the first byte is AA which is the SYNC
case study. Figure 7(a) shows the house demand and supply byte and followed by 01 which means it is a data frame. The
forecast, Figure 7(b) shows the electricity day-ahead price,and value of the phasor V1 is 2.3 kV, the angel is 0.726 rad, and
Figure 7(c) shows the amount of power that should be bought, PMU connection tester receives 24.9 frames/sec.
sold, charged, and the state of charge of the EV-Battery. As
IV. C ONCLUSION AND F UTURE W ORK
can be seen in the figure, the EV should be charged in the
early morning, when the electricity price is low and it has the In this paper we introduced our new simulation framework
required SOC level (9 kWh) before 8 AM. Indeed the charging SGsim which has been designed to simulate and investigate
occurs in the following periods , 0 to 1 and 3 to 7. In the period new approaches of smart grid applications. The framework
1 to 3 there was no charging because the electricity price was supports different types of applications such as
not cheap. • applications that require only power flow
• applications that require detailed voltage and phase mea-
C. Wide Area Monitoring surements
In this case study we show how it is possible to integrate • applications that require real time data communication
standard software into the SGsim simulator. In this example • optimization based applications
we use PMU connection tester from openPDC to capture data • application that need to deal with standards
send by a phasor measurement unit (PMU). We placed a PMU We relied on open source tools, therefore, it is possible to
at a bus in the IEEE 13 test feeder. Using the OpenDSS, introduce new features inside the framework. With three case
 Fig. 8. Snapshot of PMU connection tester when data arrives from a PMU

studies in the field of economical dispatch and optimal power [11] D. Chassin, K. Schneider, and C. Gerkensmeyer, “GridLAB-D: An
flow, demand response and electrical vehicle, and wide area open-source power systems modeling and simulation environment,” in
Transmission and Distribution Exposition Conference. Chicago, IL:
monitoring, we have shown the wide range of applications the IEEE, April 2008, pp. 1–5.
presented co-simulation framework is suitable for. [12] R. D. Zimmerman, C. E. Murillo-Sánchez, and R. J. Thomas, “MAT-
As future work, we are planning to add new components POWER Steady-State Operations, Planning and Analysis Tools for
Power Systems Research and Education,” IEEE Transactions on Power
and standards as well as to add new case studies. Additionally, Systems, vol. 26, no. 1, pp. 12–19, 2011.
we are planning to provide the framework as open source for [13] “GridSpice,” http://code.google.com/p/gridspice/, 2014, [Online; ac-
education and research purposes. cessed 20-January-2014].
[14] “Open Phasor Data Concentrator (OpenPDC),” http://openpdc.codeplex.
com/.
R EFERENCES [15] “IEEE Standard for Synchrophasors for Power Systems, IEEE Std.
C37.1182005 (Revision of IEEE Std. 13441995).”
[16] “INET Framework,” http://inet.omnetpp.org, 2014, [Online; accessed 20-
[1] “Integrated Simulation Envirement Language (INSEL) ,” January-2014].
http://www.insel.eu/, 2014, [Online; accessed 20-January-2014]. [17] “MiXiM,” http://mixim.sourceforge.net/, 2014, [Online; accessed 20-
[2] “i7-AnyEnergy ,” www7.cs.fau.de/energy/, 2014, [Online; accessed 20- January-2014].
January-2014]. [18] “Mobility Framework for OMNeT++,” http://mobility-
[3] P. Bazan and R. German, “Hybrid Simulation of Renewable Energy fw.sourceforge.net/, 2014, [Online; accessed 20-January-2014].
Generation and Storage Grids,” in Winter Simulation Conference, Berlin, [19] OpenDSS Manual, EPRI Electrical Power Research Institute, 2012.
Germany, December 2012. [20] S. G. Johnson, “The NLopt nonlinear-optimization package,” http://ab-
[4] “Xj technologies company ltd,” http://www.xjtek.com/, 2013, [Online; initio.mit.edu/nlopt, 2014, [Online; accessed 20-January-2014].
accessed 20-January-2013]. [21] M. Berkelaar et al., “lpSolve: interface to Lp solve
[5] “EnergyPLAN ,” http://www.energyplan.eu/, 2014, [Online; accessed v. 5.5 to solve linear/integer programs,” http://cran.r-
20-January-2014]. project.org/web/packages/lpSolve/lpSolve.pdf, 2013, [Online; accessed
[6] “Times ,” http://www.iea-etsap.org/web/Times.asp, 2014, [Online; ac- 20-January-2014].
cessed 20-January-2014]. [22] “MathWorks,” http://www.mathworks.de/, 2014, [Online; accessed 20-
[7] “The General Algebraic Modeling System (GAMS),” January-2014].
http://www.gams.com, 2014, [Online; accessed 20-January-2014]. [23] “The R Project for Statistical Computing,” http://www.r-project.org/,
[8] M. Lévesque, D. Qian Xu, G. Joós, and M. Maier, “Communications 2014, [Online; accessed 20-January-2014].
and power distribution network co-simulation for multidisciplinary smart
grid experimentations,” in ANSS ’12 Proceedings of the 45th Annual
Simulation Symposium, Orlando, Florida, April 2012.
[9] A. Varga, “The OMNeT++ Discrete Event Simulation System,” in Euro-
pean Simulation Multiconference (ESM 2001), Prague, Czech Republic,
June 2001.
[10] “EPRI Electrical Power Research Institute,”
http://sourceforge.net/projects/electricdss/, 2013, [Online; accessed
5-August-2013].

View publication stats