Control systems play an important role in current industries and machines, most of which still use old, hardwired devices. The industry is reluctant to use wireless technology because the communication in control systems is inefficient and easy to hack. However, recent research has shown that so-called "event-based control systems" can generate chaotic patterns in a very efficient way. This project will investigate how these chaotic controllers can be systematically designed to realize a hacker-proof and efficient wireless control system.

-

Computer models play an essential role in modern society. Learning models from data is not sustainable due to the ever-growing requirement of computational power and training data. I will develop groundbreaking methods that will learning model from data much faster and with a much smaller carbon footprint than currently needed.

In this research project, we will maximize energy efficiency of offshore wind power generation, the most important renewable energy source for The Netherlands, via a novel online learning and optimization framework and a ``semi-decentralized' control architecture. The results will be validated in a control-oriented low-fidelity parametric model and tested in a high-fidelity wind farm numerical simulator.

Real-world systems are challenged by constantly increasing complexity and uncertainty. The underlying uncertainty models are often unknown and dynamically varying and may need to be described locally with sufficient detail. In this proposal, I will optimize the geometry and time evolution of such uncertain uncertainty and explore its implications in the reliable deployment of autonomous systems.