275003 research outputs found
Sort by
Design Optimization of Dual Active Bridge Converter for Supercapacitor Application
No full text1154411557Supercapacitor (SC) is an energy storage suitable for meeting short-term requirements in power conversion systems. However, the low and variable terminal voltage of SC-based energy storage poses challenges to the design of its power electronic interface (PEI) to achieve a high round-trip efficiency and a rapid response of the energy storage system. The PEI design methodologies existing in the literature do not clearly tackle the mitigation of switching and conduction losses over the entire operating range of the SC. This article proposes a design strategy for an SC-interface dual active bridge (DAB) converter. The terminal characteristics of the SC are incorporated into an analytical design formulation aiming to maximize the DAB efficiency over the SC discharging cycle. The resulting optimization problem, addressing both conduction and switching losses, is solved numerically to obtain the optimal circuit parameters. The realization of the design objectives and the obtained efficiency is validated in circuit simulation and experiments on a 250 W laboratory prototype with a 125 V dc bus and a nominal SC voltage of 37.5 V, illustrating the improvement achieved over the conventional design approach.39
Toward video-rate quantum ghost imaging
No full textQuantum Ghost Imaging (QGI) is a powerful imaging technique that enables probing of an object using illumination levels beyond classical limits and does not rely on a single-photon-sensitive camera in the spectrum of interest. Current "heralded" QGI setups provide high-resolution images with intensified charge-coupled device (ICCD) cameras, but their acquisition time and applicability are limited by setup complexity and detector dead time. Recently, new setups using single photon detection and time-tagging have been shown to allow more efficient acquisition while also enabling new applications, such as remote 3D imaging using "asynchronous" QGI. Here, we demonstrate novel results of two asynchronous QGI setups, using a dedicated high duty-cycle single photon avalanche diode array to drastically reduce acquisition time to sub-second regime, demonstrating video acquisition at 10 fps. As this scheme allows interoperability with arbitrary single photon timing detectors, it can be adapted to a variety of applications and is not bound by the detection window of silicon-based detectors. We further study the impact of the choice of bucket detector and pump laser, using readily available off-the-shelf detectors and lasers. Summarizing the findings, we discuss the remaining limitations for real-time imaging and give an outlook on upcoming developments and an outline of further applications of both detectors and detection scheme.101
Improvement of Reflectance Spectroscopy for Oxide Layers on 4H-SiC
No full text109114In this work, we investigate the use of reflectance spectroscopy as an accurate, fast, and non-destructive method for measuring the thickness of transparent layers, such as SiO2, with thicknesses below 200 nm for microelectronic applications. To this end, we fabricated different oxides and analyzed their reflectance spectra using reflectance spectroscopy. The results were compared to theoretical reflectance spectra to validate the method. We introduce key factors to ensure accurate measurement by modeling the reflectance spectra of thin oxide layers with thicknesses ≥ 15 nm on 4H-SiC using the transfer matrix method (TMM).35
Prototyping und Validierung intelligenter Sensorik für die industrielle Zustandsüberwachung
No full text2631Beim großflächigen Einsatz von verteilter, intelligenter Sensorik im Industriekontext gelten gehobene Ansprüche an die Qualität und Zuverlässigkeit solcher Systeme vor einem Rollout in großen Stückzahlen. Der gesamte Prozess vom Prototyp bis zum Serienprodukt muss durch eine innovative Test- und
Validierungsstrategie begleitet werden, die das jeweilige Sensorsystem für die spätere Einsatzumgebung so realitätsnah wie möglich qualifiziert. So können spätere Nacharbeiten oder Ausfallrisiken minimiert werden, die oft mit hohen Kosten einhergehen. Der Artikel gibt eine Übersicht über einen möglichen Validierungsprozess am Beispiel eines prototypischen Sensorsystems Tribok der Firma Coderitter. Konkrete, im Anwendungskontext erhobene Messdaten verdeutlichen die Vorteile des beschriebenen Verfahrens
Exploring the Relationship Between Network Similarity and Transferability of Adversarial Attacks
No full text15071512Neural networks are vulnerable to adversarial attacks, and several defenses have been proposed. Designing a robust network is a challenging task given the wide range of attacks that have been developed. Therefore, we aim to provide insight into the influence of network similarity on the success rate of transferred adversarial attacks. Network designers can then compare their new network with existing ones to estimate its vulnerability. To achieve this, we investigate the complex relationship between network similarity and the success rate of transferred adversarial attacks. We applied the Centered Kernel Alignment (CKA) network similarity score and used various methods to find a correlation between a large number of Convolutional Neural Networks (CNNs) and adversarial attacks. Network similarity was found to be moderate across different CNN architectures, with more complex models such as DenseNet showing lower similarity scores due to their architectural complexity. Layer similarity was highest for consistent, basic layers, while specialized layers showed greater variability. Adversarial attack success rates were generally consistent for non-transferred attacks, but varied significantly for some transferred attacks, with complex networks being more vulnerable. We found that a DecisionTreeRegressor can predict the success rate of transferred attacks for all black-box and Carlini & Wagner attacks with an accuracy of over 90%, suggesting that predictive models may be viable under certain conditions. However, the variability of results across different data subsets underscores the complexity of these relationships and suggests that further research is needed to generalize these findings across different attack scenarios and network architectures
Enhancing Customer Shopping Experience Through AR Mini-Games
No full text1628In this research article we introduce an Augmented Reality (AR) application specifically designed to use in supermarkets. Our primary goal is to incorporate interactive AR mini-games that bring joy and digital engagement to customers during their visit. Additionally, we aim to investigate the impact of this campaign on brand awareness for both the featured goods and the supermarkets themselves. This paper presents the findings of a week-long user study conducted in two physical stores of a major retailer, involving a total of 431 participants. We analyse the impact of gameplay location and the game type (generic or product-specific) on the overall user experience. Furthermore, we investigate the impact of playing the mini-games on brand recall, which did not yield significant results during our observations. We found that our gamification approach for supermarket visits is particularly effective in attracting new customers rather than retaining existing ones
Ångström-scale surface metrology enabled by a compact milliwatt-class HHG source
No full textIn this contribution, we demonstrate high-resolution reflection imaging in the extreme ultraviolet spectral range, achieving sub-nanometer scale axial resolution and, at the same time, high imaging throughput enabled by a milliwatt-class HHG source
Sensor-based Stride Segmentation and Gait Parameter Extraction Using a Hidden Markov Model in Patients with Hereditary Spastic Paraplegia
No full text4752Hereditary Spastic Paraplegia (HSP) comprises a group of neurodegenerative disorders causing progressive lower limb symptoms primarily affecting gait functions. Progressive symmetric gait spasticity poses challenges for gait analysis. Traditional motion capture systems offer precise gait parameters but are confined to laboratory settings. In contrast, wearable sensor technology allows for gait analysis settings in real-world environments. Yet, accurate segmentation of gait cycles and ex-traction of temporal parameters remain challenging, particularly in the context of HSP. This paper introduces a comprehensive approach for gait analysis in patients with HSP, comprising two main components: stride segmentation utilizing a Hidden Markov Model (HMM), enabling robust identification of gait cycles from sensor data, and event detection and temporal gait parameter extraction. By leveraging the inherent temporal dynamics of gait patterns captured by wearable sensors, our method aims to overcome the limitations of existing techniques and provide reliable insights into the gait characteristics of HSP patients. Validation of this approach reveals promising results, with an F1 score of 89% for segmentation achieved through to-fold cross-validation. Additionally, our method demonstrates a mean absolute error of 0.008 seconds for stride time estimation compared to a gold standard motion capture system indicating the validity of our approach and its potential utility in hospital and real-world settings
Optimizing Radio Resources for Radar Services in ISAC Systems by Deep Reinforcement Learning
No full textIn integrated sensing and communication (ISAC) systems, the radar and communication functionality share the same infrastructure and radio resources. The flexible access scheme in mobile communication systems allows for variable and efficient radio resource allocation for multiple users and services. In this paper, we present a resource allocation strategy for Orthogonal Frequency Division Multiple (OFDM) based radar sensing. Furthermore, we propose a Deep Reinforcement Learning (DRL) method combined with classic OFDM radar signal processing techniques to optimize the radio resources for radar sensing. Two agents are trained with the DRL method on simulated data to predict the radio resources for the subsequent signal, aiming to achieve the required radar performance while improving the resource efficiency. One agent optimizes the transmission power, while the other optimizes the signal bandwidth/duration. The investigated scenario is a highway with traffic. We evaluate the performance of the agents based on detection loss and radio resource efficiency the metrics. Further, we compare the results against random and maximum resource-selecting agents
Encapsulating In Vitro Transcribed circRNA into Lipid Nanoparticles Via Microfluidic Mixing
No full text247260This chapter serves as a guide for researchers embarking on circular RNA-based translational studies. It provides a foundation for the successful encapsulation of circular RNA into lipid nanoparticles (LNPs) and facilitates progress in this emerging field. Crucial scientific methods and techniques involved in the formulation process, particle characterization, and downstream processing of circ-LNPs are covered. The production of in vitro transcribed circular RNA-containing LNPs based on a commercially available lipid mix is provided, in addition to the fundamentals for successful encapsulation based on lipid mixes composed of single components. Furthermore, the transfection and validation protocols for the identification of a functional and potentially therapeutic circRNA candidate for initial in vitro verification, before subsequent LNP studies, are explained.276