LoTuS Expertensystem
| dc.contributor.author | Elserafi, Ghada | |
| dc.contributor.author | Fürst, Raoul | |
| dc.contributor.author | Kleemann, Jonathan | |
| dc.contributor.author | Öztas, Batuhan | |
| dc.date.accessioned | 2024-10-20T09:01:06Z | |
| dc.date.available | 2024-10-20T09:01:06Z | |
| dc.date.created | 2024 | |
| dc.date.issued | 2024-10-20 | |
| dc.description | The LoTuS research project has developed a pilot cleaning system as a research and demonstration facility. Accompanying this, the LoTuS Expert System ("LoTuS Expertensystem") has been created as an additional user interface, serving the following functions: - Machine operation in "research mode" with flexible process configuration - Configuration and control of an additional process regulation - Process monitoring using retrofitted measurement technology and energy monitoring. This dataset includes the the code for the main functions of the expert system, as well as the training data and validation results of the individual tools developed for monitoring and controlling the drying process. | de_DE |
| dc.identifier.uri | https://tudatalib.ulb.tu-darmstadt.de/handle/tudatalib/4370 | |
| dc.language.iso | en | de_DE |
| dc.rights.license | CC-BY-4.0 (https://creativecommons.org/licenses/by/4.0) | |
| dc.subject | Parts cleaning | de_DE |
| dc.subject | Parts drying | de_DE |
| dc.subject | Virtual sensor | de_DE |
| dc.subject | Expert system | de_DE |
| dc.subject | Computer vision | de_DE |
| dc.subject | Thermography | de_DE |
| dc.subject | Fuzzy logic control | de_DE |
| dc.subject | Machine learning | de_DE |
| dc.subject.classification | 4.11-06 | |
| dc.subject.ddc | 670 | |
| dc.title | LoTuS Expertensystem | de_DE |
| dc.type | Dataset | de_DE |
| dc.type | Software | de_DE |
| dc.type | Image | de_DE |
| dc.type | Model | de_DE |
| dc.type | Other | de_DE |
| dcterms.accessRights | openAccess | |
| person.identifier.orcid | #PLACEHOLDER_PARENT_METADATA_VALUE# | |
| person.identifier.orcid | #PLACEHOLDER_PARENT_METADATA_VALUE# | |
| person.identifier.orcid | #PLACEHOLDER_PARENT_METADATA_VALUE# | |
| person.identifier.orcid | #PLACEHOLDER_PARENT_METADATA_VALUE# | |
| tuda.history.classification | Version=2020-2024;401-06 Produktionsautomatisierung | |
| tuda.unit | TUDa |
Files
Original bundle
1 - 4 of 4
| Name | Description | Size | Format | |
|---|---|---|---|---|
| Virtual_Sensor.zip | This folder contains relevant data, code and results relating to the low-cost virtual sensor developed to predict part dryness (d) from drying process data. The virtual sensor is based on a machine learning (ML) model developed following CRISP-ML(Q). | 203.22 KB | ZIP-Archivdateien | |
| Framework_functions.ipynb | The following Jupyter notebook is a compilation of the different functions used within the "LoTuS Expertensystem" framework for integrating the process monitoring and control functions. The process monitoring function contains components for importing and assessing the part dryness from the developed virtual sensor and computer vision models. Furthermore, it conveys the results to the process control function, which in terms operates in a cascade of sub-controllers to infer control actions from the developed fuzzy inference system. The control actions are subsequently passed to the machine SPS via OPC UA to adjust the corresponding actuator setpoint. The notebook contains several hard-coded values for operation initiation as well as default values in case no user input is given. | 21.86 KB | Unknown data format | |
| Fuzzy_Control.zip | This dataset includes the data and code associated with the developed fuzzy logic controller (FLC) for needs-based parts drying. The data is retrieved from field tests conducted on a throughput cleaning machine at the ETA research factory. | 763.07 KB | ZIP-Archivdateien | |
| IR-Computer_Vision.zip | This dataset contains code, images and configuration data associated with the developed thermography-based evaluation of part dryness. The analysis is performed in multiple analysis steps, including semantic segmentation and object detection via a trained YOLOv8 model. | 26.17 MB | ZIP-Archivdateien |