Universität Stuttgart

Now showing 1 - 3 of 3

Open Access
Combining knowledge and information - graph-based description of driving scenarios to enable holistic vehicle safety
(2023) Bechler, Florian; Fehr, Jörg; Neininger, Fabian; Knöß, Stefan; Grotz, Bernhard
Currently, vehicle safety is based on knowledge from injury values, crash pulses, and driving kinematics which leads to intervention strategies separated into isolated domains of active and passive safety. In this contribution, it is shown how vehicle safety can be approached holistically, allowing for human-centered and scenario-based safety decision-making. For this purpose, information from interior and exterior vehicle sensors can be linked by a mathematical framework, combining the knowledge that is already available in the individual domains. A universal graph representation for driving scenarios is developed to master the complexity of driving scenarios and allow for an optimized and scenario-based intervention strategy to minimize occupant injury values. This novel approach allows for the inclusion of sub-models, expert knowledge, results from previous simulations, and annotated databases. The resulting graph can be expanded dynamically for other objects or occupants to reflect all available information to be considered in case of urgency. As input, interior and exterior vehicle sensor data is used. Further information about the driving situation is subsequently derived from this input and the interaction between those states is described by the graph dynamically. For example, occupant attentiveness is derived from measurable eye gaze and eyelid position. From this quantity, reaction time can be estimated in turn. Combined with exterior information, it is possible to decide on the intervention strategy like e.g. alerting the driver. Physical or data-based functional dependencies can be used to represent such interactions. The uncertainties of the inputs and from the surrogate models are included in the graph to ensure a reliable decision-making process. An example of the decision-making process, by modeling the states and actuators as partially observable Markov decision process (POMDP), shows how to optimize the airbag efficiency by influencing the head position prior to an impact. This approach can be extended by additional parameters like driving environment, occupant occupancy, and seating positions in further iterations to optimize the intervention strategy for occupants. The proposed framework integrates scenario-based driving dynamics and existing knowledge from so far separated safety systems with individual activation logic and trigger points to enable holistic vehicle safety intervention strategies for the first time. It lays the foundation to consider new safety hardware, sensor information, and safety functions through a modular, and holistic approach.
Open Access
Improved a posteriori error bounds for reduced port-Hamiltonian systems
(2024) Rettberg, Johannes; Wittwar, Dominik; Buchfink, Patrick; Herkert, Robin; Fehr, Jörg; Haasdonk, Bernard
Projection-based model order reduction of dynamical systems usually introduces an error between the high-fidelity model and its counterpart of lower dimension. This unknown error can be bounded by residual-based methods, which are typically known to be highly pessimistic in the sense of largely overestimating the true error. This work applies two improved error bounding techniques, namely (a) a hierarchical error bound and (b) an error bound based on an auxiliary linear problem , to the case of port-Hamiltonian systems. The approaches rely on a secondary approximation of (a) the dynamical system and (b) the error system. In this paper, these methods are adapted to port-Hamiltonian systems. The mathematical relationship between the two methods is discussed both theoretically and numerically. The effectiveness of the described methods is demonstrated using a challenging three-dimensional port-Hamiltonian model of a classical guitar with fluid–structure interaction.
Open Access
Low-dimensional data-based surrogate model of a continuum-mechanical musculoskeletal system based on non-intrusive model order reduction
(2023) Kneifl, Jonas; Rosin, David; Avci, Okan; Röhrle, Oliver; Fehr, Jörg
Over the last decades, computer modeling has evolved from a supporting tool for engineering prototype design to an ubiquitous instrument in non-traditional fields such as medical rehabilitation. This area comes with unique challenges, e.g. the complex modeling of soft tissue or the analysis of musculoskeletal systems. Conventional modeling approaches like the finite element (FE) method are computationally costly when dealing with such models, limiting their usability for real-time simulation or deployment on low-end hardware, if the model at hand cannot be simplified without losing its expressiveness. Non-traditional approaches such as surrogate modeling using data-driven model order reduction are used to make complex high-fidelity models more widely available regardless. They often involve a dimensionality reduction step, in which the high-dimensional system state is transformed onto a low-dimensional subspace or manifold, and a regression approach to capture the reduced system behavior. While most publications focus on one dimensionality reduction, such as principal component analysis (PCA) (linear) or autoencoder (nonlinear), we consider and compare PCA, kernel PCA, autoencoders, as well as variational autoencoders for the approximation of a continuum-mechanical system. In detail, we demonstrate the benefits of the surrogate modeling approach on a complex musculoskeletal system of a human upper-arm with severe nonlinearities and physiological geometry. We consider both, the model’s deformation and the internal stress as the two main quantities of interest in a FE context. By doing so we are able to create computationally low-cost surrogate models which capture the system behavior with high approximation quality and fast evaluations.

Universität Stuttgart

Browse

Filters

Settings

Sort By

Results per page

Search Results