Fraunhofer Institute for Wind Energy Systems
Fordatis - Forschungsdaten-Repositorium der Fraunhofer-GesellschaftNot a member yet
228 research outputs found
Sort by
Probability maps and search strategies for automated UAV search in the Wadden Sea: simulation data.
No full textThe simulation framework in this dataset is intended for use in research related to search and rescue operations. It contains terrain-based movement simulation of a lost person in the Wadden Sea, probability map generation methods, search patterns and search metrics for a UAV
Unter Druck: Nachahmung von Druckgeschwüren mit in vitro Vollhautmodellen
No full textCollected research data (measurements) for in vitro full-thickness skin models from three biological donors A, B, C, which were injured with pressure-induced wounds employing a magnet. The data tables list the results of metabolic analyses (glucose, lactate), viability markers (lactate dehydrogenase), as well as reactive oxygen species (ROS) and analysis of protein levels over the various durations of the pressure treatment and subsequent reperfusion phase, which support the figures in the publication.Our work was funded by the Central Innovation Program (ZIM) by the German Federal Ministry for Economic Affairs and Energy (BMWi) with grant agreement number ZF 4353102NK7. We acknowledge the collaboration with the Prisman GmbH and thank especially Dr. Kathrin Benzig for her scientific supportcsv data tables (to be read with Microsoft Excel or any text editor
Inverse Bestimmung der Jiles-Atherton-Parameter durch Optimierung (Differential-Evolution) und Surrogatmodellierung
No full textThe dataset includes a COMSOL model that replicates the experimental setup used for characterizing ferromagnetic hysteresis. Based on this model, synthetic training data were generated and used to create a surrogate model within COMSOL. This surrogate model was integrated into a MATLAB script that also processes experimental measurement data. The script performs an inverse identification of the Jiles–Atherton parameters using a Differential Evolution algorithm. This approach enables a model-based reconstruction of the ferromagnetic hysteresis curve for the investigated material.COMSOL Multiphysics and MATLAB are required to open and execute the provided files
PackWISE: Datensatz zur Instanzsegmentierung von Verpackungsabfällen für Training und Evaluation von Modellen
No full textThe dataset comprises 586 RGB images of post-consumer lightweight packaging waste captured on a conveyor belt. Annotations follow the COCO format. The samples were kindly provided by a recycling plant of Lobbe RSW GmbH in Iserlohn, Germany. Samples were recorded as received, with adherent contaminants and without pretreatment.
This dataset is part of the multi-sensor-dataset of the project K3I-Cycling and funded by the Federal Ministry of Research, Technology and Space (BMFTR), Germany under the funding reference 033KI201.This dataset is part of the multi-sensor-dataset of the project K3I-Cycling and funded by the Federal Ministry of Research, Technology and Space (BMFTR), Germany under the funding reference 033KI201.Images are provided as JPEG files of size 4096x4096 pixel. Annotations follow the COCO format: each instance is delineated by a pixel-level mask, a bounding-box, and assigned to one of 23 categories. Bounding box coordinates are given as Integers. Pixel masks are binary run-length encoded (RLE) and can be decoded using pycocotools. Annotations can be converted to YOLO format (bounding boxes only) using the python script convert_dataset_to_yolo_format.py in the root folder of the dataset.
The dataset is divided into three splits: training (70%), validation (15%), and test (15%).
Image acquisition employed a prism-based RGB line-scan camera (SW-4000T-10GE) with a resolution of 4096 pixels. The line rate was set to 1278 lines per second. Two-dimensional images were reconstructed by scanning the samples as they moved along a conveyor belt at 0.2 m/s. Illumination was provided by 12 halogen lamps
Comparative Analysis and Numerical Stability Assessment of Jiles-Atherton Model Variants (Source Code Release)
No full textThis repository provides the full source code and datasets used to compare multiple Jiles–Atherton (JA) hysteresis model variants, including Jiles86, Jiles92, Jiles94, Bergqvist96, Annakkage00, Cheng18, and Xue22. The material contains the numerical framework for computing complete B(H) cycles using a Runge–Kutta integration scheme, the stability assessment across 300 consecutive cycles, and the RMSE-based evaluation of convergence behavior under soft- and hard-magnetic parameter sets. The code enables full reproducibility of the model comparison, including parameter handling, cycle generation, and visualization tools.This dataset contains the numerical implementation and results of a systematic comparison of multiple Jiles–Atherton (JA) hysteresis model variants. For a fixed JA parameter set, B(H) curves were computed for each model variant using a fourth-order Runge–Kutta integration scheme in MATLAB. The simulations were carried out for a range of excitation amplitudes of the magnetic field .
For each model, the numerical convergence behaviour was evaluated by calculating the root-mean-square error (RMSE) between consecutive B(H) cycles over 300 full cycles. In a subsequent step, cross-model deviations were quantified by computing the RMSE between the converged B(H) curves of the different JA model variants. The resulting dataset provides a reproducible basis for assessing numerical stability, model deviation, and sensitivity to excitation amplitude across established JA formulations
Dataset
No full textThese datasets accompany the WES preprint: Watson, Wolken-Möhlmann & Gottschall (2025), Evaluating the Impact of Motion Compensation on Turbulence Intensity Measurements from Continuous-Wave and Pulsed Floating Lidars (DOI: 10.5194/wes-2025-45).
They cover the FINO3 campaign (German North Sea; 10-min aggregation; wind sector 220°–300°) and compare raw vs motion-compensated TI from a floating ZX300M (cw) and a floating Windcube V2.1 (pulsed) together with a fixed ZX300M reference at 101 m, against Met Mast Cup TI from 101 m.File 1 — Binned statistics (public)
Title: Binned TI statistics at 101 m (raw vs motion-compensated) with fixed lidar reference
What’s inside: Wind-speed–binned (0.5 m s⁻¹) TI metrics from the analysis period 2024-04-06 to 2024-07-09: mean, min, max, std, Q90, count, MBE/RMBE, RMSE/RRMSE, representative TI error.
Instruments: Floating ZX300M (cw), floating Windcube V2.1 (pulsed), fixed ZX300M.
Columns: WLBZ6_Raw, WLBZ6_Compensated_TI, WLBW4_Raw, WLBW4_Compensated_TI, FINO3_Fixed_Lidar.
Processing: Deterministic motion compensation for LoS tilt/rotation and platform surge/sway/heave/tilt.File 2 - 10-minute time series data
Title: TI time series at 101 m (raw vs motion-compensated) with fixed lidar reference
What’s inside: Underlying 10-min TI used for File 1.
Instruments: Floating ZX300M (cw), floating Windcube V2.1 (pulsed), fixed ZX300M.
Columns: WLBZ6_Raw_TI_101_m, WLBZ6_Compensated_TI_101_m, WLBW4_Raw_TI_101_m, WLBW4_Compensated_TI_101_m, FINO3_Fixed_Lidar_TI_101_m.
Processing: Deterministic motion compensation for LoS tilt/rotation and platform surge/sway/heave/tilt.
Access: Repository record and metadata are public and citable. Files are temporarily restricted due to ongoing commercial use in a third-party Stage-3 maturity certification. Non-commercial research access may be granted on request under a mutual Data Use Agreement (DUA). Terms will be reviewed after certification completion and, if certification is not granted, annually thereafter.
How to request: Email the corresponding author ([email protected]) with institution, non-commercial purpose, intended analyses/outputs, and requested period.Python-specific binary format (.pkl) file (written using Python 3.11.9)
RF Communication Signal Dataset for Wireless Protocol Recognition based on Deep Embeddings (Part I)
No full textThis dataset supplements the paper "Approaching Domain Generalisation with Embeddings for Robust Discrimination and Recognition of RF Communication Signals" and is composed as follows:
(1) The file "synthetic.zip" contains a total of 4,000 synthetically generated signals, created from 1,000 distinct synthetic wireless protocols. For each protocol, 4 signal examples were generated, each consisting of 16,384 samples. These signals represent the training data used for the RF signal embedding models.
(2) The file "vuorenmaa.zip" contains signal bursts of various types extracted from the dataset "Radio-Frequency Control and Video Signal Recordings of Drones" (https://doi.org/10.5281/zenodo.4264467; Authors: Miika Vuorenmaa, Jaakko Marin, Mikko Heino, Matias Turunen, & Taneli Riihonen; License: CC BY 4.0 - https://creativecommons.org/licenses/by/4.0/) and is part 1/3 of the evaluation data.
(3) The file "basak.zip" contains signal bursts of various types extracted from the dataset "Drone RF Dataset" (https://doi.org/10.48804/HZRVNZ; Authors: Sanjoy Basak, Sofie Pollin & Bart Scheers; License: CC BY-NC 4.0 - https://creativecommons.org/licenses/by-nc/4.0/) and is part 2/3 of the evaluation data.
(4) The files "files_train.json" and "files_eval.json" contain the splitting of the evaluation dataset used in section "4.2. Downstream classification task" of the paper.
(5) The file "plot_spectrograms.py" shows how to access the signal data using Python and visualizes the training and evaluation data. For the script to work, the contents of all *.zip files must be unzipped and located in the same folder as the Python script and the *.json files.
NOTE: Part 3/3 of the evaluation data has been moved to a separate dataset (https://fordatis.fraunhofer.de/handle/fordatis/461) due to incompatible licences.Dieser Datensatz ergänzt das Paper „Approaching Domain Generalisation with Embeddings for Robust Discrimination and Recognition of RF Communication Signals“ und setzt sich wie folgt zusammen:
(1) Die Datei „synthetic.zip“ enthält insgesamt 4.000 synthetisch erzeugte Signale, die aus 1.000 verschiedenen synthetischen Funkprotokollen erstellt wurden. Für jedes Protokoll wurden 4 Signalbeispiele erzeugt, die jeweils aus 16.384 Samples bestehen. Diese Signale repräsentieren die Trainingsdaten, die für die RF-Signaleinbettungsmodelle verwendet werden.
(2) Die Datei „vuorenmaa.zip“ enthält Signalbursts verschiedener Typen, die aus dem Datensatz „Radio-Frequency Control and Video Signal Recordings of Drones“ (https://doi.org/10.5281/zenodo.4264467; Autoren: Miika Vuorenmaa, Jaakko Marin, Mikko Heino, Matias Turunen, & Taneli Riihonen; Lizenz: CC BY 4.0 - https://creativecommons.org/licenses/by/4.0/) und ist Teil 1/3 der Evaluierungsdaten.
(3) Die Datei „basak.zip“ enthält Signalbursts verschiedener Typen, die aus dem Datensatz „Drone RF Dataset“ (https://doi.org/10.48804/HZRVNZ; Autoren: Sanjoy Basak, Sofie Pollin & Bart Scheers; Lizenz: CC BY-NC 4.0 - https://creativecommons.org/licenses/by-nc/4.0/) und ist Teil 2/3 der Evaluierungsdaten.
(4) Die Dateien „files_train.json“ und „files_eval.json“ enthalten die Aufteilung des Evaluierungsdatensatzes, der in Abschnitt "4.2. Downstream classification task" des Papers verwendet wird.
(5) Die Datei „plot_spectrograms.py“ zeigt, wie man mit Python auf die Signaldaten zugreift und visualisiert die Trainings- und Evaluierungsdaten. Damit das Skript funktioniert, muss der Inhalt aller *.zip Dateien extrahiert und in demselben Ordner platziert werden, in dem sich das Python-Skript sowie die *.json Dateien befinden.
ANMERKUNG: Teil 3/3 der Evaluierungsdaten wurde in einen separaten Datensatz (https://fordatis.fraunhofer.de/handle/fordatis/461) wegen Lizenz-Inkompabilitäten ausgelagert.SigMF (https://sigmf.org) is used as format for storing signal and metadata.SigMF (https://sigmf.org) wird als Format für die Speicherung von Signalen und Metadaten verwendet
Synthetic Reference Energy Community load profiles for artificial case studies
No full textThis open Data set provides a reference energy community profile set for further scenario analysis. The aim is to reuse the same power flow data set in various follow up studies to facilitate the comparison of different energy community setups.This data was generated within the Fraunhofer Cluster of excellence Integrated Energy Systems in the Vor-Ort-Systeme activit
Supplementary Data of the Conducted Experiments in the Conference Paper
No full textThis dataset includes Large Language Models (LLM) prompts and the cover text, used for the three experiments conducted in the conference paper "Security and Detectability Analysis of Unicode Text Watermarking Methods Against Large Language Models". It also includes the results of all three experiments
Unraveling processes in an AWE with reference electrodes and EIS
No full textThis item contains data from experiments