Thema der Dissertation:
Global Sensitivity Analysis of Kinetic Monte Carlo Models Using Cramér-von Mises Distance Thema der Disputation:
Mathematics in Industrial Planning: The Race Against Complexity
Global Sensitivity Analysis of Kinetic Monte Carlo Models Using Cramér-von Mises Distance Thema der Disputation:
Mathematics in Industrial Planning: The Race Against Complexity
Abstract: Part 1: Mathematics in Industrial Planning: The Race Against Complexity
In the high-stakes world of industrial planning, the "perfect" mathematical solution is often the enemy of the "on-time" schedule. While optimizers provide theoretical elegance, the computational complexity of modern factory scheduling frequently exceeds available production windows, making pure scientific methods too slow for real-world use. To meet customer expectations, we must pivot from pure optimization to pragmatic heuristics that bypass these bottlenecks. This talk explores how a "spice" of mathematics, applied to non-scientific shortcuts, allows us to deliver viable results exactly when the factory needs them—trading theoretical perfection for industrial speed.
Part 2: Dissertation Summary - Advanced Computational Frameworks for Sensitivity Analysis in Heterogeneous Catalysis
This thesis presents a novel sensitivity analysis framework designed to optimize catalytic systems by identifying influential reactions within complex networks. By coupling a new methodology based on the Cramér-von Mises (CvM) distance with kinetic Monte Carlo (kMC) simulations, we overcome the traditional limitations of variance- and gradient-based approaches. This framework provides a robust, computationally efficient solution for handling uncertainties in reaction energies and barriers, even when dealing with noisy data from Density Functional Theory (DFT) calculations. Validated through case studies on CO oxidation and $CO_2$ reduction, this work bridges the gap between atomic-scale modeling and experimental validation, offering a foundation for more reliable predictions in industrial catalysis.
In the high-stakes world of industrial planning, the "perfect" mathematical solution is often the enemy of the "on-time" schedule. While optimizers provide theoretical elegance, the computational complexity of modern factory scheduling frequently exceeds available production windows, making pure scientific methods too slow for real-world use. To meet customer expectations, we must pivot from pure optimization to pragmatic heuristics that bypass these bottlenecks. This talk explores how a "spice" of mathematics, applied to non-scientific shortcuts, allows us to deliver viable results exactly when the factory needs them—trading theoretical perfection for industrial speed.
Part 2: Dissertation Summary - Advanced Computational Frameworks for Sensitivity Analysis in Heterogeneous Catalysis
This thesis presents a novel sensitivity analysis framework designed to optimize catalytic systems by identifying influential reactions within complex networks. By coupling a new methodology based on the Cramér-von Mises (CvM) distance with kinetic Monte Carlo (kMC) simulations, we overcome the traditional limitations of variance- and gradient-based approaches. This framework provides a robust, computationally efficient solution for handling uncertainties in reaction energies and barriers, even when dealing with noisy data from Density Functional Theory (DFT) calculations. Validated through case studies on CO oxidation and $CO_2$ reduction, this work bridges the gap between atomic-scale modeling and experimental validation, offering a foundation for more reliable predictions in industrial catalysis.
Zeit & Ort
27.02.2026 | 14:30
Seminarraum 007/008
(Fachbereich Mathematik und Informatik, Arnimallee 6, 14195 Berlin)
&
WebEx