Code for PhD Thesis "Nonlinear Energy-Based Systems: Modeling, Control and Numerical Realization"

Read full text Summarise with VIRAT
Authors: Attila Karsai
Cited by: 0
References: 59

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Code for PhD Thesis "Nonlinear Energy-Based Systems: Modeling, Control and Numerical Realization" This repository contains the code for the numerical experiments in the thesis. Reproducing the results Install uv by following the installation instructions. Download the code and switch to the extraction folder. Create an empty .build directory in the root folder of the codebase, i.e., where the file readme.md is located. Optional: Download the associated dataset to use precalculated results, place in results/pickle directory. Run experiments (the plots will be placed in the results/figures directory):uv run python plots/dg_intro.pyuv run python plots/dg_qsr.pyuv run python plots/dg.pyuv run python plots/mpg_cahn_hilliard_state.pyuv run python plots/mpg_convergence.pyuv run python plots/mpg_doubly_nonlinear_parabolic_state.pyuv run python plots/mpg_energybalance.pyuv run python plots/oc_casadi.pyuv run python plots/oc_state_adjoint.pyuv run python plots/pc_passivity.pyuv run python plots/pc_performance.py Some hints Throughout the codebase, the time index is always at position `0`. The state `z` thus is stored in an array with shape `z.shape == (number_of_timepoints, dimension)`. Since the implementation uses the algorithmic differentiation capabilities of JAX, the implementations of all functions need to be written in a JAX-compatible fashion. The provided examples should be a good starting point.

Fields of Study & Subjects

Control and Stability of Dynamical Systems Numerical Solvers for Ordinary Differential Equations Stability and Controllability of Differential Equations

Publication details

Journal / source
arXiv (Cornell University)
Publication date
31 March 2026
Language
EN
Document type
Preprint