Refine
Year of publication
Document Type
- Conference Proceeding (870)
- Article (reviewed) (408)
- Article (unreviewed) (123)
- Part of a Book (64)
- Letter to Editor (28)
- Patent (28)
- Book (27)
- Doctoral Thesis (14)
- Contribution to a Periodical (13)
- Working Paper (4)
- Report (3)
- Image (1)
- Moving Images (1)
- Other (1)
- Periodical Part (1)
Conference Type
- Konferenzartikel (678)
- Konferenz-Abstract (131)
- Sonstiges (34)
- Konferenz-Poster (22)
- Konferenzband (8)
Language
- English (1586) (remove)
Has Fulltext
- no (1586) (remove)
Is part of the Bibliography
- yes (1586) (remove)
Keywords
- RoboCup (32)
- Dünnschichtchromatographie (24)
- Kommunikation (15)
- Machine Learning (15)
- Gamification (13)
- TRIZ (13)
- Export (12)
- Adsorption (10)
- Biomechanik (10)
- Energieversorgung (10)
Institute
- Fakultät Elektrotechnik und Informationstechnik (E+I) (bis 03/2019) (460)
- Fakultät Maschinenbau und Verfahrenstechnik (M+V) (445)
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (296)
- Fakultät Wirtschaft (W) (227)
- Fakultät Medien und Informationswesen (M+I) (bis 21.04.2021) (134)
- ivESK - Institut für verlässliche Embedded Systems und Kommunikationselektronik (121)
- INES - Institut für nachhaltige Energiesysteme (115)
- IMLA - Institute for Machine Learning and Analytics (67)
- ACI - Affective and Cognitive Institute (51)
- Fakultät Medien (M) (ab 22.04.2021) (35)
Open Access
- Closed Access (622)
- Open Access (537)
- Closed (237)
- Bronze (114)
- Diamond (43)
- Hybrid (10)
- Gold (9)
- Grün (5)
A two-dimensional single-phase model is developed for the steady-state and transient analysis of polymer electrolyte membrane fuel cells (PEMFC). Based on diluted and concentrated solution theories, viscous flow is introduced into a phenomenological multi-component modeling framework in the membrane. Characteristic variables related to the water uptake are discussed. A Butler–Volmer formulation of the current-overpotential relationship is developed based on an elementary mechanism of electrochemical oxygen reduction. Validated by using published V–I experiments, the model is then used to analyze the effects of operating conditions on current output and water management, especially net water transport coefficient along the channel. For a power PEMFC, the long-channel configuration is helpful for internal humidification and anode water removal, operating in counterflow mode with proper gas flow rate and humidity. In time domain, a typical transient process with closed anode is also investigated.
The state-of-the-art electrochemical impedance spectroscopy (EIS) calculations have not yet started from fully multi-dimensional modeling. For a polymer electrolyte membrane fuel cell (PEMFC) with long flow channel, the impedance plot shows a multi-arc characteristic and some impedance arcs could merge. By using a step excitation/Fourier transform algorithm, an EIS simulation is implemented for the first time based on the full 2D PEMFC model presented in the first part of this work. All the dominant transient behaviors are able to be captured. A novel methodology called ‘configuration of system dynamics’, which is suitable for any electrochemical system, is then developed to resolve the physical meaning of the impedance spectra. In addition to the high-frequency arc due to charge transfer, the Nyquist plots contain additional medium/low-frequency arcs due to mass transfer in the diffusion layers and along the channel, as well as a low-frequency arc resulting from water transport in the membrane. In some case, the impedance spectra appear partly inductive due to water transport, which demonstrates the complexity of the water management of PEMFCs and the necessity of physics-based calculations.
Synthesizing voice with the help of machine learning techniques has made rapid progress over the last years [1]. Given the current increase in using conferencing tools for online teaching, we question just how easy (i.e. needed data, hardware, skill set) it would be to create a convincing voice fake. We analyse how much training data a participant (e.g. a student) would actually need to fake another participants voice (e.g. a professor). We provide an analysis of the existing state of the art in creating voice deep fakes and align the identified as well as our own optimization techniques in the context of two different voice data sets. A user study with more than 100 participants shows how difficult it is to identify real and fake voice (on avg. only 37 percent can recognize a professor’s fake voice). From a longer-term societal perspective such voice deep fakes may lead to a disbelief by default.
Synthesizing voice with the help of machine learning techniques has made rapid progress over the last years. Given the current increase in using conferencing tools for online teaching, we question just how easy (i.e. needed data, hardware, skill set) it would be to create a convincing voice fake. We analyse how much training data a participant (e.g. a student) would actually need to fake another participants voice (e.g. a professor). We provide an analysis of the existing state of the art in creating voice deep fakes and align the identified as well as our own optimization techniques in the context of two different voice data sets. A user study with more than 100 participants shows how difficult it is to identify real and fake voice (on avg. only 37% can recognize a professor’s fake voice). From a longer-term societal perspective such voice deep fakes may lead to a disbelief by default.
This work provides a series of methane adsorption isotherms and breakthrough curves on one 5A zeolite and one activated carbon. Breakthrough curves of CH4 were obtained from dynamic column measurements at different temperature and pressure conditions for concentrations of 4.4 – 17.3 mol.‐% in H2/CH4 mixtures. A simple model was developed to simulate the curves using measured and calculated data inputs. The results show that the model predictions agree very well with the experiments.
The separation of nitrogen and methane from hydrogen-rich mixtures is systematically investigated on a recently developed binder-free zeolite 5A. For this adsorbent, the present work provides a series of experimental data on adsorption isotherms and breakthrough curves of nitrogen and methane, as well as their mixtures in hydrogen. Isotherms were measured at temperatures of 283–313 K and pressures of up to 1.0 MPa. Breakthrough curves of CH4, N2, and CH4/N2 in H2 were obtained at temperatures of 300–305 K and pressures ranging from 0.1 to 6.05 MPa with different feed concentrations. An LDF-based model was developed to predict breakthrough curves using measured and calculated data as inputs. The number of parameters and the use of correlations were restricted to focus on the importance of measured values. For the given assumptions, the results show that the model predictions agree satisfactorily with the experiments under the different operating conditions applied.
Regarding the importance of adsorptive removal of carbon monoxide from hydrogen-rich mixtures for novel applications (e.g. fuel cells), this work provides a series of experimental data on adsorption isotherms and breakthrough curves of carbon monoxide. Three recently developed 5A zeolites and one commercial activated carbon were used as adsorbents. Isotherms were measured gravimetrically at temperatures of 278–313 K and pressures up to 0.85 MPa. Breakthrough curves of CO were obtained from dynamic column measurements at temperatures of 298–301 K, pressures ranging from 0.1 MPa to ca. 6 MPa and concentrations of CO in H2/CO mixtures of 5–17.5 mol%. A simple mathematical model was developed to simulate breakthrough curves on adsorbent beds using measured and calculated data as inputs. The number of parameters and the use of correlations to evaluate them were restricted in order to focus the importance of measured values. For the given assumptions and simplifications, the results show that the model predictions agree satisfactorily with the experimental data at the different operating conditions applied.
As a basis for the evaluation of hydrogen storage by physisorption, adsorption isotherms of H2 were experimentally determined for several porous materials at 77 K and 298 K at pressures up to 15 MPa. Activated carbons and MOFs were studied as the most promising materials for this purpose. A noble focus was given on how to determine whether a material is feasible for hydrogen storage or not, dealing with an assessment method and the pitfalls and problems of determining the viability. For a quantitative evaluation of the feasibility of sorptive hydrogen storage in a general analysis, it is suggested to compare the stored amount in a theoretical tank filled with adsorbents to the amount of hydrogen stored in the same tank without adsorbents. According to our results, an “ideal” sorbent for hydrogen storage at 77 K is calculated to exhibit a specific surface area of >2580 m2 g−1 and a micropore volume of >1.58 cm3 g−1.