Refine
Year of publication
- 2018 (259) (remove)
Institute
- Fachbereich Medizintechnik und Technomathematik (68)
- Fachbereich Elektrotechnik und Informationstechnik (44)
- IfB - Institut für Bioengineering (41)
- INB - Institut für Nano- und Biotechnologien (25)
- Fachbereich Luft- und Raumfahrttechnik (24)
- Fachbereich Maschinenbau und Mechatronik (24)
- Fachbereich Chemie und Biotechnologie (22)
- Fachbereich Energietechnik (22)
- Fachbereich Wirtschaftswissenschaften (20)
- Fachbereich Bauingenieurwesen (16)
Document Type
- Article (125)
- Conference Proceeding (77)
- Part of a Book (31)
- Book (12)
- Working Paper (3)
- Conference: Meeting Abstract (2)
- Doctoral Thesis (2)
- Patent (2)
- Part of a Periodical (2)
- Conference Poster (1)
Keywords
- Datenschutz (2)
- Digitale Transformation (2)
- Energy efficiency (2)
- Engineering optimization (2)
- Literaturanalyse (2)
- MINLP (2)
- Pump System (2)
- Serious Game (2)
- Water (2)
- Agility (1)
This paper presents NLP Lean Programming
framework (NLPf), a new framework
for creating custom natural language processing
(NLP) models and pipelines by utilizing
common software development build systems.
This approach allows developers to train and
integrate domain-specific NLP pipelines into
their applications seamlessly. Additionally,
NLPf provides an annotation tool which improves
the annotation process significantly by
providing a well-designed GUI and sophisticated
way of using input devices. Due to
NLPf’s properties developers and domain experts
are able to build domain-specific NLP
applications more efficiently. NLPf is Opensource
software and available at https://
gitlab.com/schrieveslaach/NLPf.
Sleep scoring is a necessary and time-consuming task in sleep studies. In animal models (such as mice) or in humans, automating this tedious process promises to facilitate long-term studies and to promote sleep biology as a data-driven f ield. We introduce a deep neural network model that is able to predict different states of consciousness (Wake, Non-REM, REM) in mice from EEG and EMG recordings with excellent scoring results for out-of-sample data. Predictions are made on epochs of 4 seconds length, and epochs are classified as artifactfree or not. The model architecture draws on recent advances in deep learning and in convolutional neural networks research. In contrast to previous approaches towards automated sleep scoring, our model does not rely on manually defined features of the data but learns predictive features automatically. We expect deep learning models like ours to become widely applied in different fields, automating many repetitive cognitive tasks that were previously difficult to tackle.
Enzyme und Biosensorik
(2018)
Enzymbasierte Biosensoren finden seit mehr als fünf Jahrzehnten einen prosperierenden Wachstumsmarkt und werden zunehmend auch in biotechnologischen Prozessen eingesetzt. In diesem Kapitel werden, ausgehend vom Sensorbegriff und typischen Kenngrößen für Biosensoren (Abschn. 18.1), elektrochemische Enzym-Biosensoren vorgestellt und deren typischen Einsatzgebiete diskutiert (Abschn. 18.2). Ein Blick über den „Tellerrand“ hinaus zeigt alternative Transduktorprinzipien (Abschn. 18.3) und führt abschließend in aktuelle Forschungstrends ein (Abschn. 18.4).
Highly competitive markets paired with tremendous production volumes demand particularly cost efficient products. The usage of common parts and modules across product families can potentially reduce production costs. Yet, increasing commonality typically results in overdesign of individual products. Multi domain virtual prototyping enables designers to evaluate costs and technical feasibility of different single product designs at reasonable computational effort in early design phases. However, savings by platform commonality are hard to quantify and require detailed knowledge of e.g. the production process and the supply chain. Therefore, we present and evaluate a multi-objective metamodel-based optimization algorithm which enables designers to explore the trade-off between high commonality and cost optimal design of single products.
Given industrial applications, the costs for the operation and maintenance of a pump system typically far exceed its purchase price. For finding an optimal pump configuration which minimizes not only investment, but life-cycle costs, methods like Technical Operations Research which is based on Mixed-Integer Programming can be applied. However, during the planning phase, the designer is often faced with uncertain input data, e.g. future load demands can only be estimated. In this work, we deal with this uncertainty by developing a chance-constrained two-stage (CCTS) stochastic program. The design and operation of a booster station working under uncertain load demand are optimized to minimize total cost including purchase price, operation cost incurred by energy consumption and penalty cost resulting from water shortage. We find optimized system layouts using a sample average approximation (SAA) algorithm, and analyze the results for different risk levels of water shortage. By adjusting the risk level, the costs and performance range of the system can be balanced, and thus the
system’s resilience can be engineered
The Kremer-Grest (KG) bead-spring model is a near standard in Molecular Dynamic simulations of generic polymer properties. It owes its popularity to its computational efficiency, rather than its ability to represent specific polymer species and conditions. Here we investigate how to adapt the model to match the universal properties of a wide range of chemical polymers species. For this purpose we vary a single parameter originally introduced by Faller and Müller-Plathe, the chain stiffness. Examples include polystyrene, polyethylene, polypropylene, cis-polyisoprene, polydimethylsiloxane, polyethyleneoxide and styrene-butadiene rubber. We do this by matching the number of Kuhn segments per chain and the number of Kuhn segments per cubic Kuhn volume for the polymer species and for the Kremer-Grest model. We also derive mapping relations for converting KG model units back to physical units, in particular we obtain the entanglement time for the KG model as function of stiffness allowing for a time mapping. To test these relations, we generate large equilibrated well entangled polymer melts, and measure the entanglement moduli using a static primitive-path analysis of the entangled melt structure as well as by simulations of step-strain deformation of the model melts. The obtained moduli for our model polymer melts are in good agreement with the experimentally expected moduli.