arXiv subjects

Carla Feistner

Publications and source records attributed to Carla Feistner.

4 recordsLinked to original sources

arXiv2025

Time-varying sensitivity analysis for mixing in chaotic flows: a comparison study

Engineered injection and extraction systems that create chaotic advection are promising procedures for enhancing mixing between two species. Mixing efficiencies vary considerably, so carefully selecting the design parameters, like pumping rates, well locations, or operation times, is crucial. While numerous studies investigate the conditions required to achieve chaotic flow, sensitivity analyses addressing its impact on mixing have rarely been performed. However, selecting a suitable sensitivity analysis method depends on the underlying system and is often restricted by the computational cost, especially when considering complex, high-dimensional models. Moreover, the most appropriate metric to quantify mixing (e.g., plume area, peak concentration) can also be system-specific. We perform a time-varying sensitivity analysis on the mixing enhancement of two chaotic flow fields with different complexities. The rotated potential mixing (RPM) flow is parametrized using two or four hyperparameters, while the quadrupole flow utilizes 16 hyperparameters. We compare three global sensitivity analysis methods: Sobol indices, Morris scores, and a modification of the activity scores. We evaluate the temporal evolution of the sensitivity of the design parameters, compare the performance of the three methods, and highlight their potential in analyzing parameter interactions. The analysis of the RPM flow shows comparable sensitivities for all methods. Additionally, our numerical experiments show that Morris is the cheapest method, needing at most four times fewer model evaluations than Sobol to reach convergence. This motivates us to only use the computationally cheaper but as reliable Morris and activity scores on the 16-dimensional model, yielding again consistent results.

nlin.CD↗

arXiv2024

Numerical simulation and analysis of mixing enhancement due to chaotic advection using an adaptive approach for approximating the dilution index

Lagrangian particle-tracking methods are particularly suitable to study solute transport in velocity fields displaying chaotic advection. They can accurately resolve stretching and folding processes, the increase in the solute-solvent interface available for diffusion as well as Kolmogorov-Arnold-Moser (KAM) islands, non-mixing regions that limit the chaotic area in the domain and, thereby, the mixing enhancement. However, they also display limitations due to the finite number of discrete particles, particularly if we are interested in the quantification of mixing processes, which require an accurate description of the particle density or concentration gradients. In this work, we use the dilution index to quantify the temporal increase in mixing of a solute within its solvent. We introduce a new approach to select a suitable grid size for the approximation of the density function, motivated by the theory of representative elementary volumes. It preserves the central feature of the dilution index, which is monotonically increasing in time, highlighting the importance of a suitable choice for the grid size in the dilution index approximation. We use this approach to demonstrate the mixing enhancement for two chaotic injection-extraction systems that exhibit chaotic structures: a source-sink dipole and a rotated potential mixing. Using our new approach, we assess the choice of design parameters of the injection-extraction systems to effectively engineer chaotic mixing. We demonstrate the important role of diffusion in filling the KAM islands and reaching complete mixing and, consequently, the importance of avoiding numerical diffusion, which often affects Eulerian methods applied on the advection-diffusion equation.

physics.flu-dyn↗

arXiv2024

DomainLab: A modular Python package for domain generalization in deep learning

Poor generalization performance caused by distribution shifts in unseen domains often hinders the trustworthy deployment of deep neural networks. Many domain generalization techniques address this problem by adding a domain invariant regularization loss terms during training. However, there is a lack of modular software that allows users to combine the advantages of different methods with minimal effort for reproducibility. DomainLab is a modular Python package for training user specified neural networks with composable regularization loss terms. Its decoupled design allows the separation of neural networks from regularization loss construction. Hierarchical combinations of neural networks, different domain generalization methods, and associated hyperparameters, can all be specified together with other experimental setup in a single configuration file. Hierarchical combinations of neural networks, different domain generalization methods, and associated hyperparameters, can all be specified together with other experimental setup in a single configuration file. In addition, DomainLab offers powerful benchmarking functionality to evaluate the generalization performance of neural networks in out-of-distribution data. The package supports running the specified benchmark on an HPC cluster or on a standalone machine. The package is well tested with over 95 percent coverage and well documented. From the user perspective, it is closed to modification but open to extension. The package is under the MIT license, and its source code, tutorial and documentation can be found at https://github.com/marrlab/DomainLab.

cs.LG↗

arXiv2024

M-HOF-Opt: Multi-Objective Hierarchical Output Feedback Optimization via Multiplier Induced Loss Landscape Scheduling

A probabilistic graphical model is proposed, modeling the joint model parameter and multiplier evolution, with a hypervolume based likelihood, promoting multi-objective descent in structural risk minimization. We address multi-objective model parameter optimization via a surrogate single objective penalty loss with time-varying multipliers, equivalent to online scheduling of loss landscape. The multi-objective descent goal is dispatched hierarchically into a series of constraint optimization sub-problems with shrinking bounds according to Pareto dominance. The bound serves as setpoint for the low-level multiplier controller to schedule loss landscapes via output feedback of each loss term. Our method forms closed loop of model parameter dynamic, circumvents excessive memory requirements and extra computational burden of existing multi-objective deep learning methods, and is robust against controller hyperparameter variation, demonstrated on domain generalization tasks with multi-dimensional regularization losses.

cs.LG↗