Assessing the Suitability of Architectural Models for Generating Smart

Grid Co-Simulations

Markus Peter

, Dominik Vereno

, Jounes-Alexander Gross

and Christian Neureiter

Josef Ressel Centre for Dependable System-of-Systems Engineering,

Salzburg University of Applied Sciences, Urstein S

ud 1, 5412 Puch/Salzurg, Austria

{ﬁrstname.lastname}@fh-salzburg.ac.at

Keywords:

Model Based Systems Engineering, SGAM, Cyber-Physical Energy System.

Abstract:

Ensuring the reliability of critical infrastructure, such as a smart grid, is of utmost importance. The veriﬁcation

of this reliability needs to occur early in the systems engineering process. An effective method to accomplish

this veriﬁcation is to simulate a model of a smart grid. Given the complexity of such a system with diverse

subsystems, co-simulation has emerged as a leading approach due to its capability to engage various inde-

pendently developed simulators. This paper explores the interoperability between architectural models and

co-simulation. The evaluation relies on a case study implemented both as a simulation and an architectural

model, with the goal of identifying similarities and differences. The conclusion drawn is that the two tools do

not achieve full interoperability to generate a comprehensive simulation out of an architectural model. This

limitation stems from co-simulations requiring precise information at an entity level, which type-based ar-

chitectural models cannot provide. However, a proposal is put forth to use architectural models as a starting

point for generating co-simulation code skeletons. The research provides an analysis of the interoperability

challenges and suggests a practical combination of the two concepts.

1 INTRODUCTION

Current power grids face signiﬁcant problems and

limitations, prompting the development of intelligent

grids known as smart grids. These smart grids offer

a transformative approach to power distribution, re-

sponding to the complexities posed by factors such

as decentralized energy generation and the increasing

diversity of energy sources (Strasser et al., 2020).

The intricate nature of smart grids necessitates

system-level validation throughout the entire engi-

neering process (Steinbrink et al., 2017). By simu-

lating different generation and demand patterns, in-

corporating renewable energy sources, and explor-

ing grid expansion scenarios, decision-makers can as-

sess the feasibility, cost-effectiveness, and reliability

of various options. These simulations play a cru-

cial role in optimizing the design and dimensioning

of smart grid components, leading to more efﬁcient

and resilient systems. Simulation serves as a power-

ful tool to unravel the complex behaviors and interac-

https://orcid.org/0009-0002-8530-5630

https://orcid.org/0000-0002-7930-6744

https://orcid.org/0000-0002-0351-2111

https://orcid.org/0000-0001-7509-7597

tions within these systems, offering valuable insights

to stakeholders, researchers, and engineers. How-

ever, a smart grid, functioning as a system of systems

erez et al., 2013), is composed of diverse subsys-

tems. Unlike traditional systems, the components of

a smart grid are not centrally developed; instead, they

evolve in a decentralized manner with contributions

from various entities. This decentralized development

introduces complexities when attempting to construct

a comprehensive monolithic simulation, as it requires

collaboration among different organisations to create

a uniﬁed model. Co-simulation addresses this chal-

lenge by integrating simulation models of different

subsystems, allowing them to be described and solved

within their respective native environments (Palensky

et al., 2017). Importantly, this approach allows or-

ganisations to provide their simulators independently,

eliminating the need for extensive collaboration.

Despite the advantages of co-simulation, a notable

challenge arises in maintaining an overview of the in-

terconnections between different simulators. The in-

tricate nature of the smart grid and the diversity of its

components make it challenging to track how various

simulators are interconnected.

In response to this challenge, architectural mod-

Peter, M., Vereno, D., Gross, J. and Neureiter, C.

Assessing the Suitability of Architectural Models for Generating Smart Grid Co-Simulations.

DOI: 10.5220/0012739800003714

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 13th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2024), pages 38-45

ISBN: 978-989-758-702-3; ISSN: 2184-4968

els could emerge as a bridge, providing a comprehen-

sive and holistic view of the smart grid ecosystem. By

abstracting the complexities into well-deﬁned views,

architectural models offer an avenue to encapsulate

the diverse perspectives and concerns of stakeholders.

This approach has the potential to unify the heteroge-

neous simulation models provided by various stake-

holders, enhancing the overall understanding of the

system’s behavior.

The potential of this approach extends even fur-

ther when aligned with the European Guide for Power

System Testing (Strasser et al., 2020). This guide

systematically divides the engineering and validation

process of cyber-physical systems like the smart grid

into four distinct yet interconnected phases. In the de-

sign phase typically system-level requirements, appli-

cation use cases and high-level architecture speciﬁca-

tion is deﬁned. The implementation phase translates

these abstract concepts into tangible prototypes, like

hardware-in-the-loop conﬁgurations or software sim-

ulations. In the validation phase the developed pro-

totypes are subject to diverse testing. And lastly the

deployment phase marks the realization of the ﬁnal

product or application, encompassing its integration

into the operational environment and subsequent roll-

out.

Architectural models, with their capacity to pro-

vide a comprehensive and abstracted view of the

smart grid system, are well-suited for the Design

phase. In contrast, the implementation and valida-

tion phase can be effectively executed through co-

simulation, harnessing the collective capabilities of

diverse simulators to conduct thorough and compre-

hensive testing. It is also worth noting that the au-

thors underscore the current absence of an integrated

approach for engineering and validating smart grids

(Strasser et al., 2020, p. 4). Therefore, this research

aims to bridge this gap by investigating the feasibil-

ity of utilizing architectural models as a link to co-

simulation within the smart grid domain, contribut-

ing to the ﬁeld by exploring the extent to which ar-

chitectural models can support the generation of co-

simulations in the smart grid context.

This approach not only improves simulation capa-

bilities within the smart grid domain but also holds

potential for broader applications in the development

and validation of cyber-physical systems across vari-

ous domains. The insights gained could potentially be

adapted and utilized across different domains, thereby

expanding the research’s impact beyond its initial fo-

cus.

2 BACKGROUND

This chapter aims to provide a brief introduction

to the concepts of architectural models and co-

simulation to establish a shared understanding of

these crucial concepts for the research. It aims to en-

sure a common understanding of terminologies.

2.1 Architectural Models

Conceptual integrity is the most important consid-

eration in system design (Brooks, 1974). One ap-

proach to maintaining this integrity is by utilizing ar-

chitectural models with their holistic perspective and

the usage of various viewpoints on the system. Ar-

chitectural models commonly adhere to the ISO4210

concept (ISO, 2022), outlining requirements for sys-

tem, software, and enterprise architecture descrip-

tions. The standard deﬁnes how system concerns

are addressed by different viewpoints independently

of technical concepts, modeling languages or tools.

In the context of the smart grid domain, the Smart

Grid Architecture Model (SGAM) (Smart Grid Co-

ordination Group, 2012) is frequently employed in

systems engineering such as in (Uslar et al., 2019)

or (Neureiter et al., 2016a). Established through the

European Commission’s Standardization Mandate it

provides a holistic perspective on overall architecture.

This model encapsulates methodologies and view-

points concerning smart grid development, offering a

standardized breakdown of a smart grid system with

a speciﬁc emphasis on interoperability (Bruinenberg

et al., 2012). Initially designed to pinpoint gaps in

smart grid standardization (Uslar et al., 2019), SGAM

has found utility in the broader context of systems en-

gineering.

2.2 Co-Simulation

A general deﬁnition of simulation is provided by

Loper, who characterizes it as ”the execution of a

model over time”(Loper, 2015). However, according

to Gomez et al., a model is not inherently readily exe-

cutable; rather, it requires a runtime environment and

speciﬁc input trajectories (Gomes et al., 2017).

Co-simulation, as deﬁned by (Steinbrink et al.,

2017), involves the coordinated execution of two or

more models that differ not only in their represen-

tation but also in their runtime environment. In this

context, a runtime environment is a software system

facilitating model execution or solving model equa-

tions. For this paper a single pairing of a model with

its runtime environment is termed a simulation unit.

The essence of co-simulation lies in multiple sim-

Assessing the Suitability of Architectural Models for Generating Smart Grid Co-Simulations

ulation units coupled together by a software interface

(Vogt et al., 2018). This characteristic enables co-

simulation to address various aspects of complex sys-

tems, making it particularly well-suited for simulat-

ing system-of-systems scenarios, such as those found

in the smart grid (Vereno et al., 2023).

2.2.1 Co-Simulation Frameworks

The main components required for a co-simulation

framework are the Scenario, the Orchestrator, the

Simulator and the Model Instance (Barbierato et al.,

2022). In Figure 1 the structure of a general co-

simulation framework and the co-simulation frame-

work mosaik (Steinbrink et al., 2019a) which will be

used in this paper, can be seen.

has has

Co-simulation Framework

governs

Simulator

governs

Orchestator

Model Instance

link

starts

initalizes

Scenario

Mosaik

Simulator

has

governs

Scheduler/Sim

Manager

has

Model

governs

starts

link

initializes

Python Scenario

API

Figure 1: Co-simulation frameworks structure based on

(Barbierato et al., 2022).

With Model Instances beeing representations of

multiple homogeneous physicial entities, encompass-

ing a physical model that could belong to different

mathematical types, ﬁnite element methods or be-

havioural models. Simulators act as Solver, contain-

ing a speciﬁc Model Instance class. They take on

the role of communication adapters with the Orches-

trator by instantiating their Models multiple times

and overseeing the resulting collection. In practical

terms, Simulators transmit inputs received from their

counterparts via the Orchestrator and execute Orches-

trator commands on their Model Instance collection.

In reciprocation, Simulators receive outputs from the

Model Instances, sending them back to the Orchestra-

tor for further coordination. The Orchestrator over-

sees the exchange of data among the Simulators and

manages their time regulation and synchronization.

And lastly the Scenario serves as a representation of

the simulated environment and encapsulates the for-

mal knowledge of the entire cyber-physical system.

It embodies the conﬁguration provided by the co-

simulation framework, that manages the startup of the

Orchestrator, the initialization of the Simulators, and

specifying the relationships among Model Instances

(Barbierato et al., 2022).

2.3 Current State of Architecture-Based

Co-Simulation Generation

The idea of creating simulations based on architec-

tural models is not a new concept. In the automotive

ﬁeld, Binder et al. proposed an approach to explore

Industry 4.0 scenarios by implementing them as ar-

chitectural models and simulating them (Binder et al.,

2021). Similar concepts can be found in the smart

grid domain, where efforts were made to generate a

certain part of a co-simulation out of an architectural

model (Binder et al., 2020).

Another study (Binder et al., 2019) focused on de-

veloping a prototype interface within a modeling tool

to customize simulation settings. However, achieving

this required an expansion of the domain-speciﬁc lan-

guage (DSL) used for modeling the architecture. The

interface additionally facilitated the mapping of spe-

ciﬁc activity diagrams to co-simulation models.

Despite the value in exploring this topic, the cur-

rent state of the art reveals that none of the papers

demonstrated complete interoperability between co-

simulations and architectural models. The second ap-

proach even necessitated adjustments to the DSL it-

self, underscoring the need for a comprehensive un-

derstanding of the interoperability between architec-

tural models and co-simulation. Which further em-

phasizes the importance of conducting a comprehen-

sive evaluation of the interoperability between archi-

tectural models and co-simulation.

3 APPROACH

To adress this research question, general insights can

be gained through the developement of a practical

artifact connecting these concepts. For this reason

the Design-Science Research Methodology (DSRM)

(Peffers et al., 2007) was used, since it places strong

emphasis on developing a practical artifact as a cen-

tral component of the research process. DSRM is

characterized by iteration cycles, where each itera-

tion contributes to the reﬁnement of our practical ar-

tifact. With this paper being at the end of one itera-

tion. Each iteration cycle unfolds through six iterative

steps, (Vom Brocke et al., 2020) which can be mapped

to different paragraphs of this paper:

• Problem Identiﬁcation and Motivation: In this

phase, the research problem is identiﬁed, and the

value of a solution is justiﬁed, corresponding to

the introduction of this paper.

• Deﬁntion of the Objectives for a Solution: The

objectives for a solution will be a concept to auto-

SMARTGREENS 2024 - 13th International Conference on Smart Cities and Green ICT Systems

matically generate co-simulations out of architec-

tural models.

• Design and Developement: In this phase the key

artifact of the study is developed, which will be

documented in section 4 and 5.

• Demonstration: This activity intends to demon-

strate the use of the artifact to solve one or more

instances of the project. In this case it can be

a prototype which generates a co-simulation or

parts of it from architectural models.

• Evaluation: In this phase, the effectiveness of the

artifact in solving the problem is assessed, as out-

lined in Section 5.

• Communication: Finally, during this phase, all

aspects of the problem and the designed artifact

are communicated to the relevant stakeholders

through the content of this paper.

Now for the developement of such an artifact it is nec-

essary to know which information a co-simulations

requires and what information architectural models

can provide. To gather this information a case study

of a simple smart grid can be conducted and real-

ized both as an architectural model as well as a co-

simulation. By comparing the requirements and con-

siderations for both approaches, valuable insights can

be obtained concerning the generation of simulations

from architectural models.

To develop a simple case study of a smart grid,

it is essential to establish a clear deﬁnition of what

constitutes a smart grid. While different deﬁnitions

exist, this work follows the deﬁnitions provided by

(Falk and Fries, 2012) and (SmartGrids, 2006) essen-

tially deﬁning a smart grid as an energy network com-

bined with the bidirectional delivery of energy infor-

mation to enable a more controllable energy delivery

and transmission. Concluding this deﬁnition a most

basic case study can be divided into the components

of an electric grid one or more components using that

grid and an component responsible for managing a

bidirectional ﬂow of data. One such case study could

be the charging behaviour of an electric vehicle in

combination with the energy generation of a wind tur-

bine Where a Smart Meter connects these components

in a partly bidirectional manner. The scenario entails

the charging station purchasing energy from the wind

turbine with lower cost than it obtains from the energy

market. The basic structure of such can be seen in the

Figure 2.

To realize the case study and develop the artifact,

two key tools will be utilized. First, for the creation of

Energy information

Wind Turbine

Energy price

Smart Meter

Added Charge

Charging Station

Energy price

Energy Market

Added Charge

E-Car

Consumed energy

Power Grid

Supplied energy

Figure 2: Co-simulaton information ﬂow.

architectural models, the SGAM Toolbox (Neureiter

et al., 2016b) will be employed. This modeling tool

is grounded in the concept of Model-Based Systems

Engineering and closely aligns with the Smart Grid

Architecture Model (SGAM), a standardized archi-

tecture model for the smart grid domain. The SGAM

Toolbox offers a robust foundation for creating archi-

tectural models tailored to the speciﬁc needs of this

research. Secondly, for co-simulation, the study has

selected the mosaik co-simulation framework (Stein-

brink et al., 2019b). Mosaik is purpose-built for

smart grid research and stands out due to its open-

source availability and an accessible application pro-

gramming interface (API). These attributes make it an

ideal choice for integrating diverse simulation models

within a uniﬁed co-simulation environment.

4 IMPLEMENTATION

In this chapter the realisation of the case study as ar-

chitectural model in the SGAM Toolbox as well as a

mosaik co-simulation will be done.

4.1 Realisation as Architectural Model

The modeling process within the SGAM Toolbox,

closely aligned with SGAM, employs various mod-

els to establish well-deﬁned views across distinct ab-

straction levels. These viewpoints are designed to

encapsulate concerns of different stakeholders. This

concept correlates with the SGAM cube, where indi-

vidual layers can be seen as fundamental viewpoints

addressing aspects such as business, functional, infor-

mational, communication, and physical facets. Each

viewpoint accommodates a grid structure, allowing

systematic element placement. Domains along the x-

axis represent the electric conversion chain, while the

y-axis presents a hierarchical perspective of informa-

tion management. This can be seen in Figure 3.

The modeling of each layer is done with a domain-

speciﬁc language relying on UML proﬁles imple-

Assessing the Suitability of Architectural Models for Generating Smart Grid Co-Simulations

Figure 3: SGAM Framework (Bruinenberg et al., 2012,

p. 30).

mented by the toolbox. With UML being a visual

modeling language for the architecture, design and

implementation of complex software systems. And

the DSL being speciﬁcally implemented to deliver a

basis for domain-related considerations and to cre-

ate a common basis for all stakeholders (Neureiter,

2017).

The implementation of the layers can be con-

ceptually divided into two main phases: a system

analysis phase and a system architecture phase. In

the analysis phase, the goal is to provide a com-

prehensive description of the system’s external per-

spective, encompassing the Business and Function

layers. The Business Layer aims to identify all in-

volved business actors along with their speciﬁc busi-

ness goals, while the Function Layer represents func-

tions and the interrelationships between them. The

system architecture phase involves the Information-

, Communication- and Component layers. The In-

formation Layer is designed to facilitate information

exchange, the Communication Layer speciﬁcally ad-

dresses the deﬁnition of communication protocols,

and the Component Layer provides a description of

the ICT networks employed for communication. In

Figure 4, the SGAM Information Layer is depicted,

utilizing SGAM-speciﬁc components to illustrate the

exchange of information among the components in

the case study.

4.2 Realisation as Co-Simulation

Implementing a co-simulation within the mosaik

framework, as illustrated in Figure 1, involves deﬁn-

ing simulation units, each comprising a simulator

and a model. Additionally, it is necessary to outline

the Scenario, encompassing both the instantiation of

simulators and the speciﬁcation of interconnections

among speciﬁc model instances. Furthermore the pre-

existing simulator PyPower (Scherfke, 2022) is used

to implement the structure and characteristics of a

power grid. The orchestration of the co-simulation

is achieved through the co-simulation framework. In

summary, the implementation of the case study, as

outlined in Section 3, within the mosaik framework

necessitates the development of the following com-

ponents:

• A scenario ﬁle

• A simulator and a model respectively for the fol-

lowing components

1. Electric Vehicle

2. Charging station

3. Smart meter

4. Wind turbine

The components are developed using the API offered

by mosaik. This API is employed by overriding in-

herited methods from mosaik’s simulator, model, and

scenario classes. Exemplatory, in listing 1 the over-

written step method for the class ElectricVehicle is

illustrated. Within the step method, the process of

one model instance’s step is deﬁned. In this example

of the electric vehicle model, the model determines

whether it can accept the provided charge based on its

capacity and current charge level.

def step(self, loading_factor):

if self.is_full():

self.added_charge = 0

else:

self.added_charge = self.max_power *

loading_factor

self.current_charge = min(self.

capacity, self.current_charge +

self.added_charge)

def is_full(self):

return self.current_charge >= self.

capacity

Listing 1: Step method of the electric vehicle class.

4.3 Simulation Integration

Based on the implementation of the case study a

differentiation between ﬁxed syntactic and semantic

parts and necessary information to implement differ-

ent simulations can be made. A listing of the neces-

sary information needed can be seen in table 1.

The information modeled in the SGAM Toolbox

can be extracted using the add-in functionality pro-

vided by Enterprise Architect (SparxSystems, 2019).

Each element modeled in the toolbox is assigned a

speciﬁc stereotype for identiﬁcation, allowing its at-

tributes and connections to be saved for further use.

The integration of architectural model information

into the mosaik simulation follows the approach pro-

posed by Binder et al. (Binder et al., 2019), which

SMARTGREENS 2024 - 13th International Conference on Smart Cities and Green ICT Systems

Generation Transformation Distribution DER Customer Premise

Market

Enterprise

Operation

Station

Field

Process

Business

Context

Starter Use Case

Model::Wind turbine

Starter Use Case Model:

:Smart grid infrastructure

Starter Use Case Model::

Smart meter

Starter Use Case Model:

:Charging station

Starter Use Case

Model::EV

Energy price

(from SGAM Business Layer)

Added Charge

Produced Wind

Energy

«abstraction»

Information Object Flow

«abstraction»

«Information Object Flow»

Information Object Flow

«use»

Figure 4: SGAM Information Layer.

Table 1: Necessary information needed for a mosaik co-

simulation.

Mosaik component Necessary information

Simulator

Time behavior

Amount of instances

Parameter

Attributes

Power Grid topology

Voltage and cable information

Scenario

Simulation runtime

Initial values

Model Model behavior

involves using templates. However, in this iteration,

only a code sceleton was generated. This decision

was inﬂuenced by the limitation of the integrated so-

lution for depicting model behavior, which relied on

activity diagrams. These diagrams were considered

unsuitable, as explained in more detail in the Evalua-

tion section.

5 EVALUATION

As we proceeded through the necessary steps for gen-

erating co-simulations from smart grid architecture

models, we encountered two main challenges.

1. The ﬁrst challenge revolves around the mod-

eling of speciﬁc behaviors within the architectural

framework. While certain approaches tailored to

modeling tools, such as using activity diagrams, may

sufﬁce for simpler examples, it’s evident that this

method isn’t universally applicable for all behaviors.

Additionally, different simulated entities are often

modeled using a variety of tools. For instance, while

automotive-speciﬁc modeling tools are most effec-

tive for modeling charging behavior, they may fall

short in adequately capturing wind generation behav-

ior. This needed diversity in modeling tools becomes

more complex when considering various stakehold-

ers, each potentially using different tools for their

specialized functionality. Addressing this challenge

could involve implementing simulators in a standard-

ized format. One such example is the Functional

Mockup Interface (Blochwitz et al., 2011), supported

by a plethora of tools. This standardization could

facilitate seamless integration of diverse simulators,

aligning with the overarching goal of uniting inde-

pendently developed simulators within a single sim-

ulation framework.

2. The second challenge lies in reconciling the

differing perspectives between architectural models

and co-simulations. Architectural models provide a

holistic view of a system through type-based models.

In contrast, co-simulations necessitate detailed spec-

iﬁcation of each speciﬁc instances of the type based

models and their interconnections within given sce-

narios. However, such speciﬁc instance information

is frequently absent within architectural models.

While there may be targeted solutions to augment

speciﬁc tools used for modeling architectural mod-

els, such as incorporating an additional instance view-

point, the ultimate goal is not a one-size-ﬁts-all tool

solution. Instead, the emphasis is on leveraging archi-

tectural models as a means to integrate independently

Assessing the Suitability of Architectural Models for Generating Smart Grid Co-Simulations

developed simulators into a broader context. Remark-

ably, tools like Powerworld (PowerWorld, 2024), em-

ployed for power system simulations, possess the ca-

pability to export network topology data in auxiliary

ﬁle formats. Therefore, akin to the aforementioned

challenge, a potential solution could involve import-

ing the missing information to bridge the interoper-

ability gap.

Overall after taking a closer look at the interop-

erability between the architectural models and co-

simulation has been taken, the conclusion can be

drawn that a full interoperability and thus a complete

simulation generation from architectural models as

proposed by Binder et al. (Binder et al., 2019) is not

deemed to be useful. Architectural models provide a

holistic view on the general function of a smart grid in

different layers and for different stakeholders. While

co-simulations wants to simulate speciﬁc use cases

of a smart grid and therefore needs speciﬁc informa-

tion about the entity relation of the smart grid compo-

nents. Therefore instead of forcing an interoperability

of concepts with different purposes a combination of

both strengths is proposed. Architectural models with

their general view on a smart grid can offer a starting

point from which simulations of different scenarios

could be started. The idea is to generate skeleton co-

simulation projects out of architectural models. By

utilizing FMU’s and integrating them in the models

the skeleton code can provide anything but intercon-

nections between the simulation units. The structure

of such an approach can be seen in Figure 5.

Figure 5: Proposed artifact.

6 CONCLUSION

In conclusion, this paper has looked into the potential

of architectural models as a foundation for generating

co-simulations within the Smart Grid domain. To as-

sess this interoperability, a case study was designed

and implemented as both an architectural model and

a co-simulation. Additionally, an artifact connecting

these two concepts by generating co-simulation code

skeletons from architectural models was developed.

However, this research revealed two notable chal-

lenges hindering the seamless generation of co-

simulations from architectural models. Firstly, the

lack of tool connection, necessitating the speciﬁcation

of various model behaviors in different tools. Sec-

ondly, the issue of entity-level modeling, where the

necessary instance view for a simulation is not inher-

ently present in the more abstract architectural model.

Despite these challenges, our research under-

scores the importance of addressing these obstacles

to unlock the full potential of architectural models

in co-simulation generation. One promising avenue

for resolution lies in the utilization of Functional

Mockup Units to abstract model behaviors, offering a

promising pathway for further exploration and eval-

uation. Additionally, ongoing efforts, such as the

model taxonomy proposed by (Vereno et al., 2024),

distinguishing between type and instance-type mod-

els, hold promise in tackling the intricacies of entity

modeling within architectural frameworks.

In summary, the study highlights the potential

of using architectural models for generating co-

simulations in the Smart Grid domain. However,

it also emphasizes the urgent requirement for ongo-

ing research and innovation to address the challenges

we’ve identiﬁed. By overcoming these hurdles, we

can fully unlock the power of architectural models as

powerful tool for comprehending, analyzing, and en-

hancing complex systems not only within the Smart

Grid but also across various other domains.

REFERENCES

Barbierato, L., Rando Mazzarino, P., Montarolo, M., Macii,

A., Patti, E., and Bottaccioli, L. (2022). A comparison

study of co-simulation frameworks for multi-energy

systems: the scalability problem. Energy Informatics,

5(4):1–26.

Binder, C., Agic, A., Neureiter, C., and L

uder, A. (2021).

Applying model-based co-simulation on modular pro-

duction units in complex automation systems. In 2021

IEEE International Symposium on Systems Engineer-

ing (ISSE), pages 1–6. IEEE.

Binder, C., Fischinger, M., Altenhuber, L., Draxler, D., Las-

tro, G., and Neureiter, C. (2019). Enabling architec-

ture based co-simulation of complex smart grid appli-

cations. Energy Informatics, 2(1):1–19.

Binder, C., Fischinger, M., Neureiter, C., Lastro, G.,

Polanec, K., and Gross, J.-A. (2020). Towards a

tool-based approach for dynamically generating co-

simulation scenarios based on complex smart grid sys-

tem architectures. In 2020 IEEE 15th International

Conference of System of Systems Engineering (SoSE),

pages 199–204. IEEE.

Blochwitz, T., Otter, M., Arnold, M., Bausch, C., Clauß, C.,

Elmqvist, H., Junghanns, A., Mauss, J., Monteiro, M.,

SMARTGREENS 2024 - 13th International Conference on Smart Cities and Green ICT Systems

Neidhold, T., et al. (2011). The functional mockup

interface for tool independent exchange of simulation

models. In Proceedings of the 8th international Mod-

elica conference, pages 105–114. Link

oping Univer-

sity Press.

Brooks, F. P. (1974). The mythical man-month. Datama-

tion, 20(12):44–52.

Bruinenberg, J., Colton, L., Darmois, E., Dorn, J., Doyle, J.,

Elloumi, O., Englert, H., Forbes, R., Heiles, J., Her-

mans, P., et al. (2012). Cen-cenelec-etsi smart grid

coordination group smart grid reference architecture.

CEN, CENELEC, ETSI, Tech. Rep, 23:24.

Falk, R. and Fries, S. (2012). Electric vehicle charging

infrastructure security considerations and approaches.

Proc. of INTERNET, pages 58–64.

Gomes, C., Thule, C., Broman, D., Larsen, P. G., and

Vangheluwe, H. (2017). Co-simulation: State of the

art. arXiv preprint arXiv:1702.00686.

ISO (2022). Software, systems and enterprise — Architec-

ture description. Standard, International Organization

for Standardization, Geneva, CH.

Loper, M. L. (2015). Modeling and simulation in the sys-

tems engineering life cycle: core concepts and accom-

panying lectures. Springer.

Neureiter, C. (2017). A domain-speciﬁc, model driven engi-

neering approach for systems engineering in the smart

grid. MBSE4U.

Neureiter, C., Engel, D., and Uslar, M. (2016a). Domain

speciﬁc and model based systems engineering in the

smart grid as prerequesite for security by design. Elec-

tronics, 5(2).

Neureiter, C., Uslar, M., Engel, D., and Lastro, G. (2016b).

A standards-based approach for domain speciﬁc mod-

elling of smart grid system architectures. In 2016 11th

System of Systems Engineering Conference (SoSE),

pages 1–6. IEEE.

Palensky, P., Van Der Meer, A. A., Lopez, C. D., Joseph,

A., and Pan, K. (2017). Cosimulation of intelligent

power systems: Fundamentals, software architecture,

numerics, and coupling. IEEE Industrial Electronics

Magazine, 11(1):34–50.

Peffers, K., Tuunanen, T., Rothenberger, M. A., and Chat-

terjee, S. (2007). A design science research method-

ology for information systems research. Journal of

management information systems, 24(3):45–77.

erez, J., D

ıaz, J., Garbajosa, J., Yag

ue, A., Gonzalez, E.,

and Lopez-Perea, M. (2013). Large-scale smart grids

as system of systems. In Proceedings of the First

International Workshop on Software Engineering for

Systems-of-Systems, pages 38–42.

PowerWorld (2024). Powerworld. https://www.

powerworld.com/.

Scherfke (2022). mosaik-pypower: A mosaik extension for

simulating power systems with pypower. https://pypi.

org/project/mosaik-pypower/. Accessed: 10.12.2023.

Smart Grid Coordination Group (2012). CEN-CENELEC-

ETSI Smart Grid Coordination Group Smart Grid Ref-

erence Architecture.

SmartGrids, E. (2006). Vision and strategy for europe’s

electricity networks of the future. European Commis-

sion.

SparxSystems (2019). Enterprise architect 15 im

Uberblick.

https://www.sparxsystems.de/ﬁleadmin/user upload/

pdfs/EAReviewersGuide EA-15-DE.pdf.

Steinbrink, C., Blank-Babazadeh, M., El-Ama, A., Holly,

S., L

uers, B., Nebel-Wenner, M., Ram

ırez Acosta,

R. P., Raub, T., Schwarz, J. S., Stark, S., Nieße, A.,

and Lehnhoff, S. (2019a). Cpes testing with mosaik:

Co-simulation planning, execution and analysis. Ap-

plied Sciences, 9(5).

Steinbrink, C., Blank-Babazadeh, M., El-Ama, A., Holly,

S., L

uers, B., Nebel-Wenner, M., Ram

ırez Acosta,

R. P., Raub, T., Schwarz, J. S., Stark, S., Nieße, A.,

and Lehnhoff, S. (2019b). Cpes testing with mosaik:

Co-simulation planning, execution and analysis. Ap-

plied Sciences, 9(5).

Steinbrink, C., Lehnhoff, S., Rohjans, S., Strasser, T. I.,

Widl, E., Moyo, C., Lauss, G., Lehfuss, F., Faschang,

M., Palensky, P., van der Meer, A. A., Heussen,

K., Gehrke, O., Guillo-Sansano, E., Syed, M. H.,

Emhemed, A., Brandl, R., Nguyen, V. H., Khavari,

A., Tran, Q. T., Kotsampopoulos, P., Hatziargyriou,

N., Akroud, N., Rikos, E., and Degefa, M. Z. (2017).

Simulation-based validation of smart grids – status

quo and future research trends. In Ma

ık, V., Wahlster,

W., Strasser, T., and Kadera, P., editors, Industrial Ap-

plications of Holonic and Multi-Agent Systems, pages

171–185, Cham. Springer International Publishing.

Strasser, T. I., de Jong, E. C., and Sosnina, M. (2020). Eu-

ropean guide to power system testing: the ERIGrid

holistic approach for evaluating complex smart grid

conﬁgurations. Springer Nature.

Uslar, M., Rohjans, S., Neureiter, C., Pr

ostl Andr

en, F.,

Velasquez, J., Steinbrink, C., Efthymiou, V., Migli-

avacca, G., Horsmanheimo, S., Brunner, H., and

Strasser, T. (2019). Applying the smart grid archi-

tecture model for designing and validating system-of-

systems in the power and energy domain: A european

perspective. Energies, 12(2):258.

Vereno, D., Harb, J., and Neureiter, C. (2023). Paving

the way for reinforcement learning in smart grid co-

simulations. In International Conference on Software

Engineering and Formal Methods, pages 242–257.

Springer.

Vereno, D., Polanec, K., Gross, J.-A., Binder, C., and

Neureiter, C. (2024). Introducing a three-layer

model taxonomy to facilitate system-of-systems co-

simulation. In 34th Annual INCOSE International

Symposium.

Vogt, M., Marten, F., and Braun, M. (2018). A survey and

statistical analysis of smart grid co-simulations. Ap-

plied Energy, 222C.

Vom Brocke, J., Hevner, A., and Maedche, A. (2020). In-

troduction to design science research. Design science

research. Cases, pages 1–13.

Assessing the Suitability of Architectural Models for Generating Smart Grid Co-Simulations