arXiv subjects

Jonas Nitzler

Publications and source records attributed to Jonas Nitzler.

12 recordsLinked to original sources

arXiv2026

Multi-Agent Framework Leveraging Knowledge Graphs for Virtual Commissioning Models

Virtual commissioning models (VCMs) of discrete manufacturing systems are used to validate automation behavior before physical deployment, but creating and maintaining them remains labor-intensive. Relevant engineering information is distributed across programmable logic controller (PLC) engineering projects, such as Siemens TIA Portal, and kinematic simulation models, such as Siemens NX Mechatronics Concept Designer (NX MCD), where it is stored in incompatible, tool-specific data structures. In practice, IEC 61131-3-based PLC programs and variables are engineered separately from rigid-body and kinematic simulation objects such as parts, joints, sensors, and actuators. As a result, understanding system behavior, generating simulation components, and mapping PLC variables to corresponding simulation objects require cross-domain expertise and remain largely manual. This paper presents a knowledge-graph-grounded multi-agent framework for semi-automated VCM development. A deterministic setup process extracts structured data from Siemens TIA Portal and Siemens NX MCD and transforms both sources into graph-based representations within a shared graph database. The framework uses a hierarchical multi-agent architecture to support three task classes in early-stage VCM development: system understanding, simulation component generation, and cross-domain signal mapping. It provides grounded natural-language access to engineering knowledge, template-guided generation of executable NX Open journal scripts, and ranked mapping suggestions between PLC variables and NX MCD simulation objects. Evaluation on a laboratory-scale discrete manufacturing system shows that the approach reduces manual cross-domain interpretation effort and makes recurring VCM engineering tasks more actionable.

cs.CE↗

arXiv2026

Scalable High-Dimensional Bayesian Field Reconstruction with Finite Elements: Application to 3D Porous Media Flow

We present a unified, finite-element-native variational inference framework for very high-dimensional Bayesian spatial field reconstruction in physics-based problems governed by partial differential equations (PDEs) that are nonlinear in the inferred parameters. The framework delivers a full-covariance Gaussian variational posterior, with a probabilistic treatment of all prior and likelihood hyperparameters, on a three-dimensional curved finite-element discretization at a stochastic field dimension exceeding 400000. To our knowledge, this is the first full-covariance variational reconstruction at this scale, complementing the low-rank Hessian-Laplace approaches that dominate extreme-scale Bayesian inversion. The spatial prior is derived from the stochastic PDE (SPDE) connection and formulated natively in terms of finite-element (FE) operators. The sparse Gaussian variational distribution is parameterized via its precision Cholesky factor, with the sparsity pattern inherited from the domain's Laplacian. Unlike covariance-based sparse parameterizations, which encode only short-range correlations, the sparse precision implicitly represents dense posterior covariances through its sparse inverse, yielding smooth, physically plausible samples at O(n) memory cost and enabling direct evidence-lower-bound (ELBO) gradients via the path-derivative (sticking-the-landing) estimator. Natural gradient strategies stabilize convergence, while a variational Bayes expectation-maximization (VB-EM) loop marginalizes all hyperparameters analytically and induces an automatic coarse-to-fine continuation. The framework is demonstrated on Bayesian permeability field reconstruction for a porous-media flow problem, recovering all major spatial features with high fidelity. Algorithmic ablation and comparison with alternative inference methods quantify the improvements over state-of-the-art baselines.

cs.CE↗

arXiv2025

Multi-Physics-Enhanced Bayesian Inverse Analysis: Information Gain from Additional Fields

Inverse analysis, such as model calibration, often suffers from a lack of informative data in complex real-world scenarios. The standard remedy, designing new experimental setups, is often costly and time-consuming, while readily available but seemingly useless data are ignored. This work proposes incorporating such data from additional physical fields into the inverse analysis, even when the forward model solves a single-physics problem. A Bayesian framework easily incorporates the additional data and quantifies the resulting uncertainty reduction. We formally introduce the proposed method, which we denote as multi-physics-enhanced Bayesian inverse analysis. Moreover, this work is the first to quantify the reduction in parameter uncertainty by comparing the information gain from the prior to the posterior when using single-physics versus multi-physics data. We demonstrate the potential of the proposed method in two exemplary applications. Our results show that even a few or noisy data points from an additional physical field can considerably increase the information gain, even when the physical field is only weakly or one-way coupled. Overall, this work proposes and promotes the future use of multi-physics-enhanced Bayesian inverse analysis as a cost- and time-saving game-changer across various fields of science and industry, particularly in medicine.

cs.CE↗

arXiv2025

QUEENS: An Open-Source Python Framework for Solver-Independent Analyses of Large-Scale Computational Models

A growing challenge in research and industrial engineering applications is the need for repeated, systematic analysis of large-scale computational models, for example, patient-specific digital twins of diseased human organs: The analysis requires efficient implementation, data, resource management, and parallelization, possibly on distributed systems. To tackle these challenges and save many researchers from annoying, time-consuming tasks, we present QUEENS (Quantification of Uncertain Effects in Engineering Systems), an open-source Python framework for composing and managing simulation analyses with arbitrary (physics-based) solvers on distributed computing infrastructures. Besides simulation management capabilities, QUEENS offers a comprehensive collection of efficiently implemented state-of-the-art algorithms ranging from routines for convergence studies and common optimization algorithms to more advanced sampling algorithms for uncertainty quantification and Bayesian inverse analysis. Additionally, we provide our latest cutting-edge research in multi-fidelity uncertainty quantification, efficient multi-fidelity Bayesian inverse analysis, and probabilistic machine learning. QUEENS adopts a Bayesian, probabilistic mindset but equally supports standard deterministic analysis without requiring prior knowledge of probability theory. The modular architecture allows rapid switching between common types of analyses and facilitates building sophisticated hierarchical algorithms. Encouraging natural incremental steps and scaling towards complexity allows researchers to consider the big picture while building towards it through smaller, manageable steps. The open-source repository is available at https://github.com/queens-py/queens.

cs.CE↗

arXiv2025

Efficient Bayesian multi-fidelity inverse analysis for expensive and non-differentiable physics-based simulations in high stochastic dimensions

High-dimensional Bayesian inverse analysis (dim >> 100) is mostly unfeasible for computationally demanding, nonlinear physics-based high-fidelity (HF) models. Usually, the use of more efficient gradient-based inference schemes is impeded if the multi-physics models are provided by complex legacy codes. Adjoint-based derivatives are either exceedingly cumbersome to derive or non-existent for practically relevant large-scale nonlinear and coupled multi-physics problems. Similarly, holistic automated differentiation w.r.t. primary variables of multi-physics codes is usually not yet an option and requires extensive code restructuring if not considered from the outset in the software design. This absence of differentiability further exacerbates the already present computational challenges. To overcome the existing limitations, we propose a novel inference approach called Bayesian multi-fidelity inverse analysis (BMFIA), which leverages simpler and computationally cheaper lower-fidelity (LF) models that are designed to provide model derivatives. BMFIA learns a simple, probabilistic dependence of the LF and HF models, which is then employed in an altered likelihood formulation to statistically correct the inaccurate LF response. From a Bayesian viewpoint, this dependence represents a multi-fidelity conditional density (discriminative model). We demonstrate how this multi-fidelity conditional density can be learned robustly in the small data regime from only a few HF and LF simulations (50 to 300), which would not be sufficient for naive surrogate approaches. The formulation is fully differentiable and allows the flexible design of a wide range of LF models.

cs.CE↗

arXiv2024

Bayesian Windkessel calibration using optimized 0D surrogate models

Boundary condition (BC) calibration to assimilate clinical measurements is an essential step in any subject-specific simulation of cardiovascular fluid dynamics. Bayesian calibration approaches have successfully quantified the uncertainties inherent in identified parameters. Yet, routinely estimating the posterior distribution for all BC parameters in 3D simulations has been unattainable due to the infeasible computational demand. We propose an efficient method to identify Windkessel parameter posteriors using results from a single high-fidelity three-dimensional (3D) model evaluation. We only evaluate the 3D model once for an initial choice of BCs and use the result to create a highly accurate zero-dimensional (0D) surrogate. We then perform Sequential Monte Carlo (SMC) using the optimized 0D model to derive the high-dimensional Windkessel BC posterior distribution. We validate this approach in a publicly available dataset of N=72 subject-specific vascular models. We found that optimizing 0D models to match 3D data a priori lowered their median approximation error by nearly one order of magnitude. In a subset of models, we confirm that the optimized 0D models still generalize to a wide range of BCs. Finally, we present the high-dimensional Windkessel parameter posterior for different measured signal-to-noise ratios in a vascular model using SMC. We further validate that the 0D-derived posterior is a good approximation of the 3D posterior. The minimal computational demand of our method using a single 3D simulation, combined with the open-source nature of all software and data used in this work, will increase access and efficiency of Bayesian Windkessel calibration in cardiovascular fluid dynamics simulations.

cs.CE↗

arXiv2023

Solving Bayesian Inverse Problems With Expensive Likelihoods Using Constrained Gaussian Processes and Active Learning

Solving inverse problems using Bayesian methods can become prohibitively expensive when likelihood evaluations involve complex and large scale numerical models. A common approach to circumvent this issue is to approximate the forward model or the likelihood function with a surrogate model. But also there, due to limited computational resources, only a few training points are available in many practically relevant cases. Thus, it can be advantageous to model the additional uncertainties of the surrogate in order to incorporate the epistemic uncertainty due to limited data. In this paper, we develop a novel approach to approximate the log likelihood by a constrained Gaussian process based on prior knowledge about its boundedness. This improves the accuracy of the surrogate approximation without increasing the number of training samples. Additionally, we introduce a formulation to integrate the epistemic uncertainty due to limited training points into the posterior density approximation. This is combined with a state of the art active learning strategy for selecting training points, which allows to approximate posterior densities in higher dimensions very efficiently. We demonstrate the fast convergence of our approach for a benchmark problem and infer a random field that is discretized by 30 parameters using only about 1000 model evaluations. In a practically relevant example, the parameters of a reduced lung model are calibrated based on flow observations over time and voltage measurements from a coupled electrical impedance tomography simulation.

cs.CE↗

arXiv2022

A Generalized Probabilistic Learning Approach for Multi-Fidelity Uncertainty Propagation in Complex Physical Simulations

Two of the most significant challenges in uncertainty quantification pertain to the high computational cost for simulating complex physical models and the high dimension of the random inputs. In applications of practical interest, both of these problems are encountered, and standard methods either fail or are not feasible. To overcome the current limitations, we present a generalized formulation of a Bayesian multi-fidelity Monte-Carlo (BMFMC) framework that can exploit lower-fidelity model versions in a small data regime. The goal of our analysis is an efficient and accurate estimation of the complete probabilistic response for high-fidelity models. BMFMC circumvents the curse of dimensionality by learning the relationship between the outputs of a reference high-fidelity model and potentially several lower-fidelity models. While the continuous formulation is mathematically exact and independent of the low-fidelity model's accuracy, we address challenges associated with the small data regime (i.e., only a small number of 50 to 300 high-fidelity model runs can be performed). Specifically, we complement the formulation with a set of informative input features at no extra cost. Despite the inaccurate and noisy information that some low-fidelity models provide, we demonstrate that accurate and certifiable estimates for the quantities of interest can be obtained for uncertainty quantification problems in high stochastic dimensions, with significantly fewer high-fidelity model runs than state-of-the-art methods for uncertainty quantification. We illustrate our approach by applying it to challenging numerical examples such as Navier-Stokes flow simulations and fluid-structure interaction problems.

cs.CE↗

arXiv2022

Global sensitivity analysis based on Gaussian-process metamodelling for complex biomechanical problems

Biomechanical models often need to describe very complex systems, organs or diseases, and hence also include a large number of parameters. One of the attractive features of physics-based models is that in those models (most) parameters have a clear physical meaning. Nevertheless, the determination of these parameters is often very elaborate and costly and shows a large scatter within the population. Hence, it is essential to identify the most important parameter for a particular problem at hand. In order to distinguish parameters which have a significant influence on a specific model output from non-influential parameters, we use sensitivity analysis, in particular the Sobol method as a global variance-based method. However, the Sobol method requires a large number of model evaluations, which is prohibitive for computationally expensive models. We therefore employ Gaussian processes as a metamodel for the underlying full model. Metamodelling introduces further uncertainty, which we also quantify. We demonstrate the approach by applying it to two different problems: nanoparticle-mediated drug delivery in a multiphase tumour-growth model, and arterial growth and remodelling. Even relatively small numbers of evaluations of the full model suffice to identify the influential parameters in both cases and to separate them from non-influential parameters. The approach also allows the quantification of higher-order interaction effects. We thus show that a variance-based global sensitivity analysis is feasible for computationally expensive biomechanical models. Different aspects of sensitivity analysis are covered including a transparent declaration of the uncertainties involved in the estimation process. Such a global sensitivity analysis not only helps to massively reduce costs for experimental determination of parameters but is also highly beneficial for inverse analysis of such complex models.

cs.CE↗

arXiv2022

Bayesian calibration of coupled computational mechanics models under uncertainty based on interface deformation

Calibration or parameter identification is used with computational mechanics models related to observed data of the modeled process to find model parameters such that good similarity between model prediction and observation is achieved. We present a Bayesian calibration approach for surface coupled problems in computational mechanics based on measured deformation of an interface when no displacement data of material points is available. The interpretation of such a calibration problem as a statistical inference problem, in contrast to deterministic model calibration, is computationally more robust and allows the analyst to find a posterior distribution over possible solutions rather than a single point estimate. The proposed framework also enables the consideration of unavoidable uncertainties that are present in every experiment and are expected to play an important role in the model calibration process. To mitigate the computational costs of expensive forward model evaluations, we propose to learn the log-likelihood function from a controllable amount of parallel simulation runs using Gaussian process regression. We introduce and specifically study the effect of three different discrepancy measures for deformed interfaces between reference data and simulation. We show that a statistically based discrepancy measure results in the most expressive posterior distribution. We further apply the approach to numerical examples in higher model parameter dimensions and interpret the resulting posterior under uncertainty. In the examples, we investigate coupled multi-physics models of fluid-structure interaction effects in biofilms and find that the model parameters affect the results in a coupled manner.

cs.CE↗

arXiv2021

Physics-Based Modeling and Predictive Simulation of Powder Bed Fusion Additive Manufacturing Across Length Scales

Powder bed fusion additive manufacturing (PBFAM) of metals has the potential to enable new paradigms of product design, manufacturing and supply chains while accelerating the realization of new technologies in the medical, aerospace, and other industries. Currently, wider adoption of PBFAM is held back by difficulty in part qualification, high production costs and low production rates, as extensive process tuning, post-processing, and inspection are required before a final part can be produced and deployed. Physics-based modeling and predictive simulation of PBFAM offers the potential to advance fundamental understanding of physical mechanisms that initiate process instabilities and cause defects. In turn, these insights can help link process and feedstock parameters with resulting part and material properties, thereby predicting optimal processing conditions and inspiring the development of improved processing hardware, strategies and materials. This work presents recent developments of our research team in the modeling of metal PBFAM processes spanning length scales, namely mesoscale powder modeling, mesoscale melt pool modeling, macroscale thermo-solid-mechanical modeling and microstructure modeling. Ongoing work in experimental validation of these models is also summarized. In conclusion, we discuss the interplay of these individual submodels within an integrated overall modeling approach, along with future research directions.

cs.CE↗

arXiv2021

A Novel Physics-Based and Data-Supported Microstructure Model for Part-Scale Simulation of Laser Powder Bed Fusion of Ti-6Al-4V

The elasto-plastic material behavior, material strength and failure modes of metals fabricated by additive manufacturing technologies are significantly determined by the underlying process-specific microstructure evolution. In this work a novel physics-based and data-supported phenomenological microstructure model for Ti-6Al-4V is proposed that is suitable for the part-scale simulation of selective laser melting processes. The model predicts spatially homogenized phase fractions of the most relevant microstructural species, namely the stable $β$-phase, the stable $α_{\text{s}}$-phase as well as the metastable Martensite $α_{\text{m}}$-phase, in a physically consistent manner. In particular, the modeled microstructure evolution, in form of diffusion-based and non-diffusional transformations, is a pure consequence of energy and mobility competitions among the different species, without the need for heuristic transformation criteria as often applied in existing models. The mathematically consistent formulation of the evolution equations in rate form renders the model suitable for the practically relevant scenario of temperature- or time-dependent diffusion coefficients, arbitrary temperature profiles, and multiple coexisting phases. Due to its physically motivated foundation, the proposed model requires only a minimal number of free parameters, which are determined in an inverse identification process considering a broad experimental data basis in form of time-temperature transformation diagrams. Subsequently, the predictive ability of the model is demonstrated by means of continuous cooling transformation diagrams, showing that experimentally observed characteristics such as critical cooling rates emerge naturally from the proposed microstructure model, instead of being enforced as heuristic transformation criteria.

cs.CE↗