The DOI displayed above is for this specific version of this dataset, which is currently the latest. Newer versions may be published in the future. For a link that will always point to the latest version, please use
DOI: 10.4121/e763b413-6bc0-4548-a3b4-bc83d8ad211b

He, Kanghui (2025): Data underlying the PhD thesis: Learning-based control under constraints: Towards safety and computational efficiency. Version 1. 4TU.ResearchData. dataset. https://doi.org/10.4121/e763b413-6bc0-4548-a3b4-bc83d8ad211b.v1

Categories

• Mechanical Engineering
• Electrical and Electronic Engineering
• Engineering

Keywords

Deep Reinforcement Learning (DRL) Learning-based control Model predictive control Safety-critical systems systems and control theory

Licence

CC BY-NC-ND 4.0

Interoperability

• RO-Crate Metadata

Export as...

RefWorks BibTeX Reference Manager Endnote DataCite NLM DC CFF

by Kanghui He

This dataset supports the PhD thesis titled "Learning-Based Control Under Constraints: Towards Safety and Computational Efficiency". The thesis comprises six main chapters (Chapters 2–7), and the data is organized accordingly. All simulations were conducted in a MATLAB environment.

• Chapters 2–3 focus on constrained approximate dynamic programming (ADP) algorithms for safe control. In particular, Chapter 2 presents a convex piecewise quadratic optimization algorithm designed to efficiently solve the ADP problem. Chapter 3 uses penalty methods to deal with constraints.
• Chapter 4 includes code for the explicit approximation of safety filters.
• Chapter 5 implements all-element predictive safety filters to ensure the safe control of piecewise affine (PWA) systems.
• Chapters 6–7 propose an integrated optimization- and learning-based control framework for general constrained nonlinear systems.

History

• 2025-08-08 first online, published, posted

Publisher

4TU.ResearchData

Format

mat/m/fig

Associated peer-reviewed publication

Efficient and Safe Learning-based Control of Piecewise Affine Systems Using Optimization-Free Safety Filters

Funding

• Clarinet (grant code 101018826) ERC advanced grant

Organizations

TU Delft, Faculty of Mechanical Engineering, Delft Center for Systems and Control