Refine
Document Type
Keywords
- Automatisierungstechnik (1)
- Bildklassifikation (1)
- Digitale Datenübertragung (1)
- Driving simulator, attention, mental workload, sound, physiologic sensors, electroencephalogram, virtual reality (1)
- Frequenzumrichter (1)
- GStreamer (1)
- H.264 (1)
- High Intensity Radiated Fields, Lightning Indirect Effects, helicopters, induced currents, numerical simulation, scheduled maintenance (1)
- Hochfrequenztechnik (1)
- ISO Norm 9283 (1)
Institute
- Fakultät für Elektrotechnik Feinwerktechnik Informationstechnik (19) (remove)
The illumination of registration plates poses challenges for lighting technology. To achieve the wide and steep angles required for homogeneous illumination of the license plate, optics are needed to refract and reflect light. In this work, registration plate lamps with different components are designed. After the construction, the concepts are tested for their limits in terms of position and manufacturability. The
entire work is carried out at the company CREAT GmbH in Ingolstadt. All developments in the field of registration plate lighting are subject to the UNECE and country-specific regulations, which must be considered in the development of the lights.[1] Some vehicles currently available on the automotive market are measured to classify and compare the different lamp systems. From the average values, the lamps are designed. Adjustments are repeatedly made to the lighting device to achieve the most
homogeneous illumination of the registration plate. As soon as the lights meet the legal requirements, the geometric limits of the lighting device are defined in relation to the plate. As there are different polymers to produce optics, these are simulated with their refractive indices and tested for their effects on the light image.
The objective of this study was to develop an experimental framework that allows research on whether the combined use of a virtual reality (VR) driving simulator, electroencephalography signals collected during driving tests and designed attentional paradigms could be applied to assess the performance of affective warning sounds for re-channelling driver’s attention in certain driving situations. Three VR driving scenarios with different stimuli and two driving tasks were designed and implemented. Acquisition of driver’s physiological data and behavioural metrics were integrated to be gathered synchronously. In addition, an experimental protocol was designed for the assessment of the impact of affective warning sounds. The technical feasibility of the framework was evaluated with the results showing that the proposed framework enables the data acquisition and analysis for a variety of warning stimuli. After successful technical deployment, a study with experimental subjects will follow to evaluate the effectiveness of affective warning sounds.
This study aims to compare two different implementations of speech emotion recognition models.The emphasis is directed towards evaluating their efficacy in capturing and characterizing dialogues
portrayed by actors within a film scene to create suitable musical intervals. The goal of the overarching research intends to derive indications to enhance the compositional process of film scores by recognizing the emotion in a particular scene. Based on established deep learning models, the study delves into the exploration of two distinct emotion classification metrics: The Six Emotion Prediction and the Valence/Arousal/Dominance Prediction. To facilitate a comparative analysis, a preliminary study an a following survey is deployed. The preliminary study confirms a significant difference in the generated MIDI data. For this reason, a survey is essential to detect the better fitting algorithm. Participants are tasked to rate the affective suitability of eight generated interval sequences to the corresponding film scenes. The Suitability is verified quantitatively using a bidirectional rating system. Both model assessments are conducted within a uniform sound design, thus ensuring unbiased conditions for evaluation. Upon a thorough examination of our extensive analysis, a preference for method A becomes increasingly evident.
Infradar-Localization: single-chip infrared- and radar-based Monte Carlo localization. This paper proposes a novel approach for indoor robot localization that leverages a fusion of information from single-chip infrared (Time-of-Flight) and radar sensors. The aim of our research is the development of a cost-effective and lightweight system that can achieve high-precision robot localization. Unlike traditional localization methods based on LiDARs or cameras, our proposed system uses single-chip infrared and radar sensors to overcome the limitations of high cost and bulky hardware. Specifically, we employ a Doppler radar-based velocity motion model for the estimation of the robot's ego-motion, eliminating the need for additional sensors such as IMU or wheel encoders. Next, we describe a hybrid sensor model for single-chip infrared and radar sensors that provides robust and accurate environmental perception with dynamic outlier removal. Finally, we integrate these components into a Monte Carlo localization framework to generate accurate real-time estimation of the robot's position and orientation. This is the first time a single-chip infrared and radar fusion based framework has been applied to robot localization, to the best of our knowledge. Through a comprehensive experimental evaluation, we demonstrate the system's high accuracy and efficiency, achieving an average localization error of 9 cm in diverse indoor environments. This remarkable performance, combined with the low-cost and lightweight nature of our proposed solution, positions it as a highly promising alternative for a wide range of applications, including robotics, smart homes, and autonomous vehicles. The significant advancements of this novel approach offer vast potential to revolutionize the field of localization, enabling more precise and cost-effective navigation systems.
Das Ziel der vorliegenden Bachelorarbeit ist die Entwicklung eines Lehrsystems, bestehend aus einer Lasteinheit. Dabei setzt sich die Lasteinheit aus einer Motor-Frequenzumrichter-Kombination zusammen.
Es ist sind Steuerprogramme zu entwickeln, die Funktionen, wie eine Lastsimulation, bieten. Zudem ist die Entwicklung einer GUI vorgesehen, die zur Beobachtung und Steuerung dient. Die Programment-
wicklung findet im TIA-Portal, in STARTER und in LabVIEW statt. Neben der Programmentwicklung ist eine abschließende Antriebsauslegung durchzuführen. Für die erfolgreiche Programmentwicklung wurde im ersten Schritt ein Programm im TIA-Portal implementiert, wobei dies nur grobe Züge des Endprogramms enthielt und für erste Tests zuständig war. Eine Reglerparametrierung wurde mithilfe des TIA-Programms und des Programms STARTER durchgeführt. Im Anschluss wurden die Programm-teile nach LabVIEW portiert und das Hauptprogramm bis zur endgültigen Funktionsfähigkeit erweitert. Die Antriebsauslegung wurde mithilfe von entsprechend großen Lasten und Temperaturaufnahmen durchgeführt. Das Ergebnis ist ein lauffähiges LabVIEW-Programm mit GUI, das Lasten simulieren und Bremsungen von gekoppelten Motoren durchführen kann. Die Bremsung wurde schließlich für die größtmögliche Leistungsentnahme und Antriebsauslegung eingesetzt. Der Antrieb kam bisher nicht zum Einsatz.
Deep Learning basierte Segmentierung und Transformation von Spike-Rauschen in der MR-Bildgebung
(2023)
Die Magnetresonanztomographie (MRT) ist ein bildgebendes Diagnoseverfahren, das auf starken Magnetfeldern und sensiblen Detektoren basiert. Es kann zu Spike-Rauschen kommen, das die Bildqualität negativ beeinflusst. In dieser Arbeit wird untersucht, ob neuronale Netzwerke das Spike-Rauschen reduzieren können, um die Bildqualität zu verbessern. Dazu wird eine Bildverarbeitungspipeline entworfen, bestehend aus den Teilbereichen Detektion, Segmentierung und Transformation. Verschiedene neuronale Netzarchitekturen werden für jeden Teilbereich ausgewählt und auf Basis verschiedener Daten trainiert. Die Ergebnisse zeigen, dass die Detektion und Segmentierung basierend auf der L1-Norm die besten Ergebnisse liefern. Es werden bestimmte Netzwerkarchitekturen ausgewählt, um die Transformation durchzuführen. Die gesamte Pipeline reduziert das Rauschen erheblich, aber weitere Tests sind erforderlich, um die Robustheit der Implementierung sicherzustellen.
Weiterentwicklung eines gasbetriebenen Wirbelschicht-Sinterreaktor für kleinste Keramikbauteile
(2023)
Diese Arbeit optimiert einen gasbetriebenen Wirbelschicht-Sinterreaktor für kleinste Keramikbauteile. Der Reaktor nutzt das Prinzip der Wirbelschicht, um die zu sinternden keramischen Bauteile in der Schwebe zu halten und ermöglicht so eine erhebliche Beschleunigung der Prozesszeit im Vergleich zu konventionellen Brennöfen. Ziel dieser Arbeit war es, den bestehenden Forschungs-reaktor zu optimieren, indem der Brenner neu ausgerichtet, ein neues konisches Brennrohr konstruiert und eine neue Vorrichtung zur Probenentnahme entwickelt wurde. Die Anlage wurde erfolgreich verbessert und ermöglicht nun den Schwebezustand der Bauteile sowie eine schnellere Aufheizung. Es wurde auch eine Vorrichtung zur Probenentnahme konstruiert und die Bauteile auf Maßhaltigkeit und Bildung von Agglomeraten untersucht. Allerdings muss das Material des Brennrohrs ersetzt werden. Insgesamt konnte gezeigt werden, dass die Anlage erfolgreich verbessert wurde und das Sintern der Bauteile nun möglich ist.
Capability of scheduled maintenance to detect electromagnetic and lightning protection degradation
(2022)
Due to the growing use of low-conductivity composite materials for airframes and skin, the shielding effectiveness of aircraft has been weakened within recent years. As a result, the importance of electromagnetic hazards and associated effects on flight safety have increased dramatically. Especially high intensity electromagnetic fields are likely to interfere with sensitive avionic systems, while lightning strikes produce high-energetic charges and currents which might cause physical damages and induced transients on onboard wirings and equipment-internal circuits. To ensure safe flight and landing conditions, additional protection measures are required to be introduced in aircraft design. However, aging and degradation effects of mentioned electromagnetic and lightning protection features can't be avoided, thus, specific scheduled maintenance actions within appropriate intervals must be defined during certification process to retain every single aircraft in a serviceable condition. As there's a lack of database to validate expected preliminary degradation and aging behavior until the first scheduled inspection, most maintenance definitions just rely on in-service and maintenance experience of similar aicraft.
Hence, within this thesis, current regulations and practical implementations of lightning and electromagnetic protection on Airbus rotorcraft shall be investigated in order to perform 3D simulations for determination of the correlation of electromagnetic protection failure and wiring-induced transients. As a final aim, maintenance improvements for minimization of adverse degradation impact on helicopter safety can be achieved.
The company Plusoptix GmbH develops and sells medical devices to measure the refraction of infants. For the implementation of these functionalities, the necessary hardware and software is integrated in the apparatus. In addition, an individual OS was developed to perform specific tasks efficiently and quickly. In order to continuously close security gaps and meet requirements, the operating system must constantly be up to date. This activity is realized by the construction of the embedded Linux with the help of the internal NBS. Nightly building of the OS enables testing of changed software code or integrated modules. However, due to the steadily growing project tree of Plusoptix GmbH, the system increasingly shows problem areas. This includes aspects of administration and information overview. In addition, new functionalities such as continuous integration are desired to ensure the greatest possible SQA. Based on these features, the master thesis includes the development of a CIS to address the emerging requirements. Additionally, all previous process cycles of the old system should be integrable. For this purpose, a catalog of requirements for the new system must first be designed and specified. With the framework conditions defined, a targeted search can then be made for software to solve the existing problems. These must be evaluated in order to find the most efficient solution approach. In the end, a system is to be established that replaces the NBS andmeets current and future requirements.
The time available to bring a new product to market is getting shorter and shorter. To keep up, virtual prototypes have become an essential part in the product development cycle. We wanted to create a configuration interface to allow for a fast and easy configuration of said prototypes. We want to evaluate how the created interface performs when used with valid or invalid data and how accurately the requirements for the software were met. As a basis for this analysis we have created an approach to develop a user interface (UI) based on the existing Vivifly application and have extended it with some functionality. We used the approach to perform a set of software tests to evaluate its functionality. The tests focus on the behaviour of the UI and its output, when the user interacts with it. This is done by either testing the UI as it was intended or with inputs, that usually would not be entered. The results of the software tests show that although a large part of the requirements have been implemented, some special cases have not yet been considered and need additional information. Out of the 22 test cases seven did encounter a problem while performing them.