Offenburg University of Applied Sciences
Hochschulschriftenserver der Hochschule OffenburgNot a member yet
6641 research outputs found
Sort by
Vorstellung einer sensorintegrierten Schraube zur Bestimmung der Gesamtzusatzspannung
No full textEs wird eine neu entwickelte sensorintegrierte Schraube vorgestellt, die in der Lage ist, sowohl Axialkraft als auch Biegemoment im Schraubenschaft zu messen. Die vorgestellte Lösung integriert vier Dehnungsmessstreifen im Schraubenschaft, ohne die mechanischen Eigenschaften der Schraube wesentlich zu verändern. Durch die vorgestellte neue Auswertemethodik der drei nichtzentrisch positionierten Dehnungsmessstreifen ist die Bestimmung der Gesamtzusatzspannung in der Schraubenverbindung möglich. Die Sensorschraube wurde hergestellt, kalibriert und in verschiedenen Versuchsaufbauten (axiale Kraftmessung, Vierpunktbiegeversuch) validiert. Die Ergebnisse zeigen, dass die Schraube die aufgebrachten Belastungen mit hoher Genauigkeit abbilden kann und sich für praxisnahe Anwendungen eignet.A newly developed sensor-integrated bolt is presented that is capable of measuring both axial force and bending moment in the bolt shaft. The solution presented integrates four strain gauges into the bolt shaft without significantly altering the mechanical properties of the bolt. The new evaluation method presented for the three non-centrically positioned strain gauges makes it possible to determine the total additional stress in the bolt connection
Data aggregation using peer to peer network and dynamic directed graph operation
No full textInteractions between sensors that avoid central control require peer-to-peer communication. Sensor nodes produce measurement data. The nodes use a peer-to-peer network and epidemic algorithms to disseminate messages among all nodes to solve a problem cooperatively. A caller node chooses its neighbor, as a called node, randomly in every round to send and receive messages. Push transmission occurs when a message is sent from the caller node to the called node, and pull transmission occurs when the caller node receives a message from the called node. Push-pull transmission involves both processes, allowing messages to be sent in both directions when using both push transmission and pull transmission.
This work introduces a scheme for the Push-Pull Sum communication protocol to exchange data to compute an aggregation function. In each round, every node sends half its sum and half its weight to itself and its neighbor uniformly at random. The called node that receives the message from its neighbors will reply to the caller node and itself with half of its sum and half of its weight divided by the number of nodes that send the message to the called node. The Push-Pull Sum protocol maintains the mass conservation property in all rounds. It converges to the correct aggregate of ground truth, as shown by the potential function that decreases in each round.
We compare the Push-Pull Sum protocol with other communication protocols, i.e. Push Sum, Pull Sum, Push, Pull, Push-Pull protocols. In addition, we also compare them with Push-Sampling, Pull-Sampling, Push-Pull-Sampling protocols which use Finite Population Correction to estimate the sample size for computing the aggregation function. The simulation results indicate that the Push Sum protocol and the Push protocol outperform all protocols in the long run with respect to message complexity. Meanwhile, the Push-Pull Sum protocol outperforms all protocols in terms of time complexity in the long run. The results indicate a trade-off between the Mean Squared Error and both message complexity and time complexity.
We introduce the directed Flip-Chain algorithm, which transforms the d-regular multidigraph into another d-regular multidigraph by swapping two adjacent vertices in the trail of directed edges. Repeatedly applying the directed Flip-Chain operation to an arbitrary d-regular multidigraph will build every d-regular multidigraph with the same probability in the limit. The operations of the directed Flip-Chain will transform any d-regular multidigraph into an expander with high probability and introduce randomness to the d-regular multidigraph, which is particularly useful in dynamic graphs. The directed Flip-Chain algorithm preserves the connectivity and regularity of d-regular multidigraph. Cheeger's inequality determines whether a directed graph is an expander-directed graph. The simulation results indicate that the directed vertex expansion and the directed edge conductance of d-regular multidigraphs converge after performing directed Flip-Chain operations.
In addition, the directed Flip-Chain operation is observed with the Push-Sum protocol and the Push-Pull Sum protocol for randomized rumor spreading on a regular directed graph. The simulation results indicate that the Push-Sum protocol and the Push-Pull Sum protocol with directed Flip-Chain operations improve the Mean Squared Error compared to the Push-Sum protocol and the Push-Pull Sum protocol without Flip-Chain operations regarding time complexity and message complexity for rumor dissemination
Push-Sum Protocol with Random Proportion for Agents in Energy Trading
No full textCommunication between prosumer and consumer agents to exchange data on energy consumed and produced is needed to estimate the average amount of energy available and determine prices in energy trading. Agents represented by nodes exchange the message using the random call model of communication protocol. We analyze the communication protocols of Push-Sum algorithm of Kempe et al., the r Push-Sum algorithm, and the random r Push-Sum algorithm for computing the approximate amount of energy data in a peer-to-peer network of agents. In the random r Push-Sum communication protocol, each agent determines the proportion value drawn uniformly at random in each round to determine the proportion of sum and weight that will be sent to the agent’s neighbor and the agent itself with respect to the mass conservation property. Each agent calculates the approximate value of the average amount of energy required and the average amount of energy provided in the peer-to-peer network. The Mean Squared Error (MSE) of the aggregation results is affected by the use of various proportions of sum and weight. These communication protocols improve MSE as the number of rounds and messages exchanged increases
Enhanced Low Frequency GPR Measurement Using Optimized Vivaldi Antennas
No full texthis work presents a low frequency ground-penetrating radar (GPR) measurement system employing optimized Vivaldi antennas, targeting high penetration depth and enhanced range resolution. The radar integrates a modified edge-slot Vivaldi antenna operational from 75 MHz with a center frequency of 537.5 MHz and a bandwidth of 925 MHz. An initial GPR survey conducted in a ground-coupled mode over an asphalt surface successfully detects a buried utility pipe at approximately 2 m depth with good resolution. Detailed simulations, measurements and experimental results demonstrate the improved low frequency performance of the antenna, making it a suitable candidate for deep subsurface sensing
Chancen und Risiken des Einsatzes von KI-generierten Inhalten in Dialogmarketingkampagnen
No full textDer Einsatz von KI-generierten Inhalten im Dialogmarketing bietet Unternehmen die Möglichkeit, Personalisierung, Effizienz und Skalierbarkeit ihrer Marketingkampagnen erheblich zu verbessern. Generative Künstliche Intelligenz (GenAI) kann maßgeschneiderte Texte, Bilder und andere Inhalte in Echtzeit erstellen, wodurch der Qualitäts-Quantitäts-Tradeoff überwunden werden kann. Gleichzeitig gibt es Herausforderungen, insbesondere hinsichtlich strategischer Integration, Datensicherheit und ethischer Implikationen wie Verzerrungen und Transparenz. Dieses Kapitel beleuchtet die Chancen und Risiken des Einsatzes von GenAI im Dialogmarketing anhand empirischer Untersuchungen und Praxisbeispiele von Burda Direct. Die Ergebnisse zeigen, dass KI-generierte Inhalte zu einer besseren Kundenansprache und höheren Interaktionsraten führen können, jedoch eine klare Kennzeichnung und verantwortungsbewusste Nutzung erforderlich sind
Enhancing the Omnichannel Retailing Customer Experience in the Pre-Purchase Phase: Evaluation and Improvement of a Digital Grocery Brochure
No full textRetailers are increasingly focusing on designing seamless omnichannel customer journeys to enhance the overall customer experience. By gathering information from various touchpoints used during the customer journey, retailers aim to create a seamless and engaging experience. One crucial touchpoint in the pre-purchase phase of grocery retailers is the digital brochure, which aims to capture customers’ attention with weekly offers. However, these digital brochures can sometimes cause irritation and fail to appeal to customers. The objective of this paper is to identify and address the irritations associated with a digital grocery brochure to optimize and enhance the customer experience during the pre-purchase phase of an omnichannel customer journey. To achieve this, a digital brochure from a German grocery retailer was evaluated using the Customer-Experience-Tracking Method. The analysis revealed five principal irritations. The findings suggest that implementing uniform pricing, reducing the number of products per page, ensuring clear and understandable product categorization, and using images of unpacked products can improve the customer experience
Exploring the Acceptance of a Decision-support Chatbot Among Executives: Insights from an Adapted Utaut Model
No full textThe study investigates executives’ acceptance of an AI decision-support chatbot in marketing and sales using an adapted UTAUT model. Survey results show that behavioral intention is mainly driven by performance expectancy—shaped by output quality, source trustworthiness, and time savings—and by effort expectancy, which increases when cognitive load is reduced and users have high computer self-efficacy. The findings extend UTAUT to executive decision-making contexts and offer practical implications for implementing effective AI-driven decision-support systems
Verfahren und Vorrichtung zur Rahmensynchronisation (DE102024111661A1)
No full textDie Erfindung betrifft ein Verfahren zur Rahmensynchronisation für ein digitales Empfangssignal, welches an einem Ausgang eines Übertragungskanals als Folge eines einem Eingang des Übertragungskanal zugeführten binären digitalen Sendesignals empfangen wird, wobei das binäre digitale Sendesignal eine zu übertragende Information in einem Datenpaket beinhaltet, welches in einem Datenabschnitt eine vorgegebene Anzahl D von Datensymbolen und in einem optionalen Synchronisationsabschnitt eine vorgegebene Anzahl S von Synchronisationssymbolen aufweist, und wobei das binäre digitale Sendesignal eine vorgegebene Symbolrate aufweist, wobei das digitale Empfangssignal durch eine Folge von digitalen Abtastwerten dargestellt wird und wobei eine Abtastfrequenz verwendet wird, die größer oder gleich der Symbolrate ist. Nach der Erfindung wird ein Sliding-Window definiert, welches eine der zeitlichen Länge eines Datenpakets oder des Datenabschnitts eines Datenpakets entsprechende zeitliche Länge aufweist. In jeder Position µ des Sliding-Window wird jedem Symbolabschnitt des Datenabschnitts und optional jedem Symbolabschnitt des Synchronisationsabschnitts des Sliding-Window ein Abtastwert zugeordnet. Für jede Position µ des Sliding-Window wird unter Verwendung einer vorgegebenen Metrik ein Positionswert &Lgr;met[µ] ermittelt, wobei hierfür zumindest die den Symbolabschnitten des Datenabschnitts zugeordneten Abtastwerte verwendet werden. Die Metrik basiert auf der Verwendung eines Modells für den Übertragungskanal, welches einen gedächtnisfreien Übertragungskanal annimmt und bei welchem das digitale Empfangssignal für die möglichen binären Werte des Sendesignals jeweils mit einer bedingten Wahrscheinlichkeitsdichtefunktion beschrieben wird. Aus dem Anfangszeitpunkt des Sliding-Window in einer Position, für welche die Positionswerte &Lgr;metein Maximum annehmen oder für welche der Positionswert &Lgr;meteinen vorgegebenen Schwellwert übersteigt, wird ein Startzeitpunkt eines Datenpakets in dem digitalen Empfangssignal ermittelt. Weiterhin betrifft die Erfindung eine Vorrichtung und ein Computerprogramm zur Realisierung des Verfahrens