fix references

MarleneBusch · MarleneBusch · commit 0dd213d2eebb · 2026-02-18T18:19:26.000+01:00
diff --git a/joss_paper/paper.md b/joss_paper/paper.md
@@ -44,13 +44,13 @@ bibliography: paper.bib
 
 `artist`is a software package for the simulation and optimization of concentrating solar power (CSP) plant operation. Solar tower power plants use an array of mirrors (heliostats), to reflect and concentrate sunlight onto a small area called the receiver. The thermal power, represented through a flux distributions in the digital twin, can be used directly as high-temperature heat in industrial processes or it can be converted into climate-neutral electricity. This Python package, `artist`, implements a fully differentiable digital twin for solar tower power plants, providing tools for data-driven power plant component modeling and optimized power plant operation. The differentiable ray tracer in `artist` simulates how light interacts with the three-dimensional scene, including environmental conditions, enabling gradient-based optimizations with raytracing. In solar tower plants imperfect hardware is inevitable and leads to individually deformed fluxes for each heliostat. Surface deformations [@Ulmer:2011] and misalignments due to inaccurate kinematic [@Sattler:2020] components are main contributors for uncertainties. To improve efficiency in a solar tower power plant, each heliostats deformed and misaligned flux needs to aim at an individual point on the receiver for an optimal combined flux density distribution. The digital twin `artist` provides functionality to reconstruct the real-world heliostat surfaces and the heliostat kinematic. The simulation includes algorithms for alignment and raytracing to create flux density predictions. Based on this the heliostat aim points can be optimized. The main functionality of `artist` is shown in \autoref{fig:flowchart}.
 
-![Features of `artist`. To create digital twins of solar tower power plants in `artist`, users are asked to provide HDF5-files containing data about the physical layout of the power plant. The HDF5 scenarios can be generated by `artist` from various data sources. `artist` unpacks these files to initiate the reconstruction, simulation and prediction. The optimized parameters, can be used directly as input to a power plant control software. To efficiently handle heliostat surfaces, `artist` contains a fully differentiable, parallelized NURBS implementation for 3D surfaces. \label{fig:flowchart}](flowchart.png)
-
 # Statement of Need
 
-In solar tower power plants, digital twins with precise simulation and reliable predictive capabilities are essential to realize fully autonomous power plant operation and a consequential reduction in costs [@Huang:2021]. While solar tower power plants may vary in their individual architectural details, their digital twins consistently rely on ray tracing. Conventional ray tracers for CSP [@Ahlbrink:2012; @SolTrace, @Tonatiuh:2018] achieve good results in simulating power plant behavior. However, they can only use ray tracing to make predictions based on the supplied data and the resulting model. From a machine learning perspective, these ray tracers are confined to forward computations. Therefore, they often require large amounts of data to function accurately. `artist` addresses this limitation with its differentiable implementation of the ray tracer and all connecting modules. Single differentiable digital twin tasks for CSP simulations have been addressed in related works, using inverse deep learning for heliostat surface modeling [@Lewen:2025] or generative neural networks for flux predictions [@Kuhl:2024]. In contrast to these black-box AI algorithms, `artist` maintains interpretability due to its physical models. In CSP technology, transparency in system behavior is critical for building trust in automated decision-making and software generated operation suggestions. The methodology in `artist` remains consistent across all optimization tasks, integrating modeling, prediction and optimization into one combined tool.
+In solar tower power plants, digital twins with precise simulation and reliable predictive capabilities are essential to realize fully autonomous power plant operation and a consequential reduction in costs [@Huang:2021]. While solar tower power plants may vary in their individual architectural details, their digital twins consistently rely on ray tracing. Conventional ray tracers for CSP [@Ahlbrink:2012; @SolTrace; @Tonatiuh:2018] achieve good results in simulating power plant behavior. However, they can only use ray tracing to make predictions based on the supplied data and the resulting model. From a machine learning perspective, these ray tracers are confined to forward computations. Therefore, they often require large amounts of data to function accurately. `artist` addresses this limitation with its differentiable implementation of the ray tracer and all connecting modules. Single differentiable digital twin tasks for CSP simulations have been addressed in related works, using inverse deep learning for heliostat surface modeling [@Lewen:2025] or generative neural networks for flux predictions [@Kuhl:2024]. In contrast to these black-box AI algorithms, `artist` maintains interpretability due to its physical models. In CSP technology, transparency in system behavior is critical for building trust in automated decision-making and software generated operation suggestions. The methodology in `artist` remains consistent across all optimization tasks, integrating modeling, prediction and optimization into one combined tool.
 This software is designed for researchers, power plant operators, developers within the CSP community or anyone else interested in the field. `artist` includes data loaders compatible with various data sources, including the open-access CSP database PAINT [@Phipps:2025]. Therefore, anyone including people who do not have direct access to an operational power plant can use `artist`. We aim to strengthen community engagement and collaboration for further research advancements by developing `artist` as an easily accessible open-source software.
 
+![Features of `artist`. To create digital twins of solar tower power plants in `artist`, users are asked to provide HDF5-files containing data about the physical layout of the power plant. The HDF5 scenarios can be generated by `artist` from various data sources. `artist` unpacks these files to initiate the reconstruction, simulation and prediction. The optimized parameters, can be used directly as input to a power plant control software. To efficiently handle heliostat surfaces, `artist` contains a fully differentiable, parallelized NURBS implementation for 3D surfaces. \label{fig:flowchart}](flowchart.png)
+
 # State of the Field
 
 On the one hand, commercially proven digital twin software solutions for solar tower power plants such as the raytracer STRAL [@Ahlbrink:2012] are non-differentiable and therefore lack the ability to reconstruct heliostat field models or to optimize power plant parameters. Instead they assume idealized hardware or provide tools for exact modeling which rely on infeasible measurements. These already commercially proven software solutions are typically proprietary and closed-source, limiting the possibilities for public contributions. On the other hand published differentiable proof-of-concept methods which improve digital twin simulation and enable optimization are partial, non-unified solutions which are not necessarily scalable. `artist` addresses these gaps by introducing a fully differentiable, open-source raytracer for CSP that integrates scalable, physics-informed AI methods into a single coherent framework. We build on the concept of differentiable raytracing for CSP by adding differentiable ray blocking. Furthermore we redefine existing proof-of-concept solutions under the unified approach of gradient descent optimization of physical models to ensure consistency across all digital twin tasks. As no comparable open-source alternative exists, we provide the basis for future contributions.
@@ -61,7 +61,7 @@ On the one hand, commercially proven digital twin software solutions for solar t
 
 # Research Impact Statement
 
-The underlying concepts of `artist` are based on previous publications, which demonstrate the potential of the differentiable ray tracing approach [@Pargmann:2024] and more generally show the potential of increasing solar tower power plant efficiency in order to provide an environmentally friendly solution to meet the globally rising demand for energy [@CSPRoadMapNREL; @Carballo:2020; @Edenhofer:2011].
+The underlying concepts of `artist` are based on previous publications, which demonstrate the potential of the differentiable ray tracing approach [@Pargmann:2024] and more generally show the potential of increasing solar tower power plant efficiency in order to provide an environmentally friendly solution to meet the globally rising demand for energy [@CSPRoadMapNREL; @Carballo:2025; @Edenhofer:2011].
 
 # AI Usage Disclosure
 
