Skip to content

Commit 52181e7

Browse files
maurerlemaurerle
committed
add joss paper
create using pandoc -s paper.md -o paper.pdf --bibliography paper.bib --citeproc
1 parent 7602914 commit 52181e7

File tree

5 files changed

+236
-0
lines changed

5 files changed

+236
-0
lines changed

‎README.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -16,6 +16,7 @@ SPDX-License-Identifier: AGPL-3.0-or-later
1616
[![](https://img.shields.io/pypi/l/assume-framework.svg)](https://pypi.org/pypi/assume-framework/)
1717
[![](https://img.shields.io/pypi/status/assume-framework.svg)](https://pypi.org/pypi/assume-framework/)
1818
[![](https://img.shields.io/readthedocs/assume)](https://assume.readthedocs.io/)
19+
[![DOI](https://joss.theoj.org/papers/a8843ad1978808dc593b16437a2a029e/status.svg)](https://joss.theoj.org/papers/a8843ad1978808dc593b16437a2a029e)
1920

2021
[![Try examples in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://github.com/assume-framework/assume/tree/main/examples/notebooks)
2122

‎codemeta.json

Lines changed: 44 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,44 @@
1+
{
2+
"@context": "https://raw.githubusercontent.com/codemeta/codemeta/master/codemeta.jsonld",
3+
"@type": "Code",
4+
"author": [
5+
{
6+
"@id": "https://orcid.org/0000-0001-8345-3889",
7+
"@type": "Person",
8+
"email": "[email protected]",
9+
"name": "Florian Maurer",
10+
"affiliation": "FH Aachen"
11+
},
12+
{
13+
"@id": "https://orcid.org/0000-0003-1897-3671",
14+
"@type": "Person",
15+
"email": "[email protected]",
16+
"name": "Nick Harder",
17+
"affiliation": "Universität Freiburg"
18+
},
19+
{
20+
"@id": "https://orcid.org/0009-0009-1389-4844",
21+
"@type": "Person",
22+
"email": "[email protected]",
23+
"name": "Kim Kira Miskiw",
24+
"affiliation": "Karlsruhe Institute of Technology"
25+
},
26+
{
27+
"@id": "https://orcid.org/0000-0002-3347-9922",
28+
"@type": "Person",
29+
"email": "[email protected]",
30+
"name": "Manish Khanra",
31+
"affiliation": "Fraunhofer ISI"
32+
}
33+
],
34+
"identifier": "https://doi.org/10.5281/zenodo.8088760",
35+
"codeRepository": "https://github.com/assume-framework/assume/",
36+
"datePublished": "2024-11-15",
37+
"dateModified": "2024-11-13",
38+
"dateCreated": "2024-11-13",
39+
"description": "ASSUME - Agent-based Simulation for Studying and Understanding Market Evolution",
40+
"keywords": "agent-based, energy market, simulation, reinforcement learning, bidding",
41+
"license": "AGPL-3.0-or-later",
42+
"title": "ASSUME",
43+
"version": "v0.4.3"
44+
}

‎codemeta.json.license

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
SPDX-FileCopyrightText: ASSUME Developers
2+
3+
SPDX-License-Identifier: AGPL-3.0-or-later

‎paper/paper.bib

Lines changed: 105 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,105 @@
1+
@Comment{
2+
SPDX-FileCopyrightText: ASSUME Developers
3+
4+
SPDX-License-Identifier: AGPL-3.0-or-later
5+
}
6+
7+
@incollection{maurerMarketAbstractionEnergy2023,
8+
title = {Market {{Abstraction}} of {{Energy Markets}} and {{Policies}} - {{Application}} in an {{Agent-Based Modeling Toolbox}}},
9+
booktitle = {Energy {{Informatics}}},
10+
author = {Maurer, Florian and Miskiw, Kim K. and Acosta, Rebeca Ramirez and Harder, Nick and Sander, Volker and Lehnhoff, Sebastian},
11+
editor = {J{\o}rgensen, Bo N{\o}rregaard and Da Silva, Luiz Carlos Pereira and Ma, Zheng},
12+
year = {2023},
13+
month = dec,
14+
volume = {14468},
15+
pages = {139--157},
16+
publisher = {Springer Nature Switzerland},
17+
address = {Cham},
18+
doi = {10.1007/978-3-031-48652-4_10},
19+
isbn = {978-3-031-48652-4},
20+
}
21+
22+
@inproceedings{maurerKnowYourTools2024,
23+
title = {Know {{Your Tools}} - {{A Comparison}} of {{Two Open Agent-Based Energy Market Models}}},
24+
booktitle = {2024 20th {{International Conference}} on the {{European Energy Market}} ({{EEM}})},
25+
author = {Maurer, Florian and Nitsch, Felix and Kochems, Johannes and Schimeczek, Christoph and Sander, Volker and Lehnhoff, Sebastian},
26+
year = {2024},
27+
month = jun,
28+
pages = {1--8},
29+
issn = {2165-4093},
30+
doi = {10.1109/EEM60825.2024.10609021},
31+
}
32+
33+
@inproceedings{miskiwExplainableDeepReinforcement2024,
34+
title = {Explainable {{Deep Reinforcement Learning}} for {{Multi-Agent Electricity Market Simulations}}},
35+
booktitle = {2024 20th {{International Conference}} on the {{European Energy Market}} ({{EEM}})},
36+
author = {Miskiw, Kim K. and Staudt, Philipp},
37+
year = {2024},
38+
month = jun,
39+
pages = {1--9},
40+
issn = {2165-4093},
41+
doi = {10.1109/EEM60825.2024.10608907},
42+
}
43+
44+
45+
@inproceedings{khanraEconomicEvaluationElectricity2024,
46+
title = {Economic {{Evaluation}} of {{Electricity}} and {{Hydrogen-Based Steel Production Pathways}}: {{Leveraging Market Dynamics}} and {{Grid Congestion Mitigation}} through {{Demand Side Flexibility}}},
47+
shorttitle = {Economic {{Evaluation}} of {{Electricity}} and {{Hydrogen-Based Steel Production Pathways}}},
48+
booktitle = {2024 20th {{International Conference}} on the {{European Energy Market}} ({{EEM}})},
49+
author = {Khanra, Manish and Klobasa, Marian and Patil, Parag and Scholz, Daniel},
50+
year = {2024},
51+
month = jun,
52+
pages = {1--6},
53+
issn = {2165-4093},
54+
doi = {10.1109/EEM60825.2024.10608890},
55+
}
56+
57+
@inproceedings{adamsBlockOrdersMatter2024,
58+
title = {Do {{Block Orders Matter}}? {{Impact}} of {{Regular Block}} and {{Linked Orders}} on {{Electricity Market Simulation Outcomes}}},
59+
shorttitle = {Do {{Block Orders Matter}}?},
60+
booktitle = {2024 20th {{International Conference}} on the {{European Energy Market}} ({{EEM}})},
61+
author = {Adams, Johanna and Harder, Nick and Weidlich, Anke},
62+
date = {2024-06},
63+
pages = {1--7},
64+
issn = {2165-4093},
65+
doi = {10.1109/EEM60825.2024.10608956},
66+
}
67+
68+
69+
@article{harderHowSatisfactoryCan2024,
70+
title = {How {{Satisfactory Can Deep Reinforcement Learning Methods Simulate Electricity Market Dynamics}}? {{Bechmarking}} via {{Bi-level Optimization}}},
71+
shorttitle = {How {{Satisfactory Can Deep Reinforcement Learning Methods Simulate Electricity Market Dynamics}}?},
72+
author = {Nick Harder and Lesia Mitridati and Farzaneh Pourahmadi and Anke Weidlich and Jalal Kazempour},
73+
year = {2024},
74+
month = oct,
75+
journal = {Energy Informatics Review},
76+
}
77+
78+
79+
@inproceedings{bublitzAgentbasedSimulationGerman2014,
80+
title = {Agent-Based {{Simulation}} of the {{German}} and {{French Wholesale Electricity Markets}}},
81+
booktitle = {Proceedings of the 6th {{International Conference}} on {{Agents}} and {{Artificial Intelligence}} - {{Volume}} 2},
82+
author = {Bublitz, Andreas and Ringler, Philipp and Genoese, Massimo and Fichtner, Wolf},
83+
year = {2014},
84+
month = mar,
85+
series = {{{ICAART}} 2014},
86+
pages = {40--49},
87+
publisher = {{SCITEPRESS - Science and Technology Publications, Lda}},
88+
address = {Setubal, PRT},
89+
doi = {10.5220/0004760000400049},
90+
urldate = {2023-05-15},
91+
isbn = {978-989-758-016-1},
92+
}
93+
94+
@article{schimeczekAMIRISAgentbasedMarket2023,
95+
title = {{{AMIRIS}}: {{Agent-based Market}} Model for the {{Investigationof Renewable}} and {{Integrated}} Energy {{Systems}}},
96+
shorttitle = {{{AMIRIS}}},
97+
author = {Schimeczek, Christoph and Nienhaus, Kristina and Frey, Ulrich and Sperber, Evelyn and Sarfarazi, Seyedfarzad and Nitsch, Felix and Kochems, Johannes and Ghazi, A. Achraf El},
98+
year = {2023},
99+
month = apr,
100+
journal = {Journal of Open Source Software},
101+
volume = {8},
102+
number = {84},
103+
issn = {2475-9066},
104+
doi = {10.21105/joss.05041},
105+
}

‎paper/paper.md

Lines changed: 83 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,83 @@
1+
<!--
2+
SPDX-FileCopyrightText: ASSUME Developers
3+
4+
SPDX-License-Identifier: AGPL-3.0-or-later
5+
-->
6+
7+
---
8+
title: 'ASSUME - Agent-based Simulation for Studying and Understanding Market Evolution'
9+
tags:
10+
- Python
11+
- agent based simulation
12+
- energy market
13+
- reinforcement learning
14+
- market simulation
15+
- simulation
16+
authors:
17+
- name: Florian Maurer
18+
orcid: 0000-0001-8345-3889
19+
corresponding: true
20+
affiliation: 1
21+
"affiliation": 1
22+
- name: Nick Harder
23+
orcid: 0000-0003-1897-3671
24+
"affiliation": 2
25+
- name: Kim Kira Miskiw
26+
orcid: 0009-0009-1389-4844
27+
affiliation: 3
28+
- name: Manish Khanra
29+
orcid: 0000-0002-3347-9922
30+
affiliation: 4
31+
32+
affiliations:
33+
- name: University of Applied Sciences Aachen, Germany
34+
index: 1
35+
- name: Inatech Universität Freiburg, Germany
36+
index: 2
37+
- name: Karlsruhe Institute of Technology, Germany
38+
index: 3
39+
- name: Fraunhofer Institute for Software and Systems Engineering, Germany
40+
index: 4
41+
date: 13 November 2024
42+
bibliography: paper.bib
43+
---
44+
45+
<!-- pandoc -s paper.md -o paper.pdf --bibliography paper.bib --csl=apa.csl --filter pandoc-citeproc --pdf-engine=xelatex-->
46+
47+
# Summary
48+
49+
**ASSUME** is an open-source toolbox for agent-based simulations of European electricity markets, with a primary focus on the German market setup. Developed as an open-source model, its primary objectives are to ensure usability and customizability for a wide range of users and use cases in the energy system modeling community.
50+
51+
A unique feature of the ASSUME toolbox is its integration of **Deep Reinforcement Learning** methods into the behavioral strategies of market agents. The model offers various predefined agent representations for both the demand and generation sides, which can be used as plug-and-play modules, simplifying the reinforcement of learning strategies. This setup enables research into new market designs and dynamics in energy markets.
52+
53+
# Statement of need
54+
55+
While different other agent-based models have been developed for the study of energy markets, such as PowerACE [@bublitzAgentbasedSimulationGerman2014] and AMIRIS [@schimeczekAMIRISAgentbasedMarket2023], the possible integration of reinforcement learning methods into the behavioral strategies of market agents is currently unique to ASSUME.
56+
This feature allows for the exploration of new market designs and dynamics in energy markets using a common open-source simulation framework.
57+
Further unique features of `ASSUME` are the extensive market abstraction which allows to define complex multi market scenarios as shown in [@maurerMarketAbstractionEnergy2023].
58+
ASSUME is designed to provide results which are easily comparable and interactable.
59+
This is set in contrast to other agent-based simulation frameworks, which are often designed to be used for a specific use case.
60+
Instead, various moving parts can be configured, such as the market abstraction, the agent properties and bidding strategies, or the reinforcement learning methods.
61+
62+
# Architecture
63+
64+
`ASSUME` builds on the open-source agent-based simulation library [mango-agents](https://pypi.org/project/mango-agents/) to model the simulation world and to give users continuous feedback during the simulation runs.
65+
Interaction with the results of the simulation is possible through the Grafana dashboards, which do have direct access to the simulation database.
66+
Additionally, writing the output to CSV files is supported for scenarios in which a database is not available.
67+
68+
New scenarios can be created by providing CSV files, accessing simulation data from a database or by using the object-oriented API, which makes it possible to integrate new classes, maket mechanisms and bidding strategies as well.
69+
The core overview of available classes and the interaction with the world is shown in \autoref{fig:architecture}.
70+
71+
![Basic Architecture of the ASSUME simulation\label{fig:architecture}](../docs/source/img/architecture.svg)
72+
73+
# Publications
74+
75+
ASSUME has been used to investigate the usage of complex order types like block bids in wholesale markets [@adamsBlockOrdersMatter2024] as well as for the integration of demand-side flexibility and redispatch markets in [@khanraEconomicEvaluationElectricity2024].
76+
Studies of applicability of reinforcement learning methods in energy markets were tackled in [@harderHowSatisfactoryCan2024], while an analysis of explainable AI methods was appliedd to ASSUME in [@miskiwExplainableDeepReinforcement2024].
77+
For better interoperability with other energy system model data, adapters to interact from ASSUME with PyPSA and AMIRIS are available and make a comparison to other renomated market simulation tools possible [@maurerKnowYourTools2024].
78+
79+
# Acknowledgements
80+
81+
Kim K. Miskiw, Nick Harder and Manish Khanra thank the German Federal Ministry for Economic Affairs and Climate Action for the funding of the ASSUME project under grant number BMWK 03EI1052A.
82+
83+
# References

0 commit comments

Comments
 (0)