 Automatic Control LaboratoryOpen OpportunitiesIn modern society, the problem of traffic congestion in densely populated cities is worsening due to the heavy daily commuting. This affects not only the commuters that waste many hours in traffic congestion every year, but it also creates inefficiencies and pollution that translate into additional societal costs. The role of service stations in the future years will gain importance due to, among other factors, the constant raise in electric vehicle sales. The presence of a service station on highway stretches highly affect the level of traffic congestion on the road. In this work, we aim at developing an Iterative Learning Control scheme that is able to maximize the positive effects that a service station has in terms of peak and overall traffic congestion reduction. We will validate the designed algorithm using real data and state-of-the-art micro-simulators.
- Automotive Engineering, Control Engineering, Systems Theory and Control
- Master Thesis, Semester Project
| In this project, we will design a distributionally robust model predictive controller for managing a multi-asset portfolio over T stages. Exploiting Wasserstein ambiguity sets to capture distributional uncertainty, we aim to robustify our controller against black swan events, i.e., unpredictable events with severe impact. - Engineering and Technology, Mathematical Sciences
- Master Thesis, Semester Project
| In this project, we aim to construct data-driven
discrepancy-based ambiguity sets and use them to provide novel reformulations for static and two-stage stochastic programs in the area of energy systems and markets. We will then validate the formulations with real-world data from industrial partners. - Engineering and Technology, Mathematical Sciences
- Master Thesis, Semester Project
| The need to ensure the efficient operation of dynamical systems of increasing complexity, such as smart buildings and interconnected power grids, has recently triggered renewed interest in “direct data-driven control” methods. This term refers to the design of controllers based only on measured data, without first identifying a model of the system –a process that can be costly and time-consuming. Here the goal is to design a controller that guarantees the closed-loop behavior not for a single plant, but rather for all the plants compatible with the available data.
Such a robust control objective has been often reformulated as a data-dependent optimization problem, via robust optimization tools, with the drawback of heavily relying on the noise model. Instead, this project explores an approach able to capture more general system priors and noise assumptions, where the direct data-driven control problem is reformulated as a saddle point problem: in simple terms, the idea is to test the controller against the worst plausible plant, given the experimental data. For which class of systems, and how, can the data-dependent saddle point formulation be solved in a reliable and efficient way? - Engineering and Technology, Information, Computing and Communication Sciences, Mathematical Sciences
- Master Thesis, Semester Project
| We want to examine further novel universal dual-port grid forming control for dc/ac power converters. The main advantage of this control scheme is the coupling of the converter’s ac and dc terminal voltages, making it applicable to any technology and any operation of renewable generation. - Engineering and Technology
- Bachelor Thesis, Master Thesis, Semester Project
|
|
