Master Thesis
Recent Submissions
-
Sizing, Cost, and Emissions of Vehicle to Grid Coupled PV at District Scale
(2022)This thesis investigates the potential of PV and V2G coupling for the city of Singapore. Using Singapore’s 55 planning areas as spatial unit an optimization model is developed to determine economically optimal PV scaling and charge/discharge strategies. Mobility flows between planning areas are assessed using a large set of mobile phone records, and local electricity demand and solar potentials are modeled using a bottomup approach based ...Master Thesis -
-
-
Evaluating foehn development in a changing climate using machine learning and investigating the impact of foehn on forest fire occurrence and severity
(2020)Foehn winds have a substantial impact on forest fires all over the world. However, two questions have remained unanswered to the foehn community over the last years. First, while accurate forecasts are viable at the timescale of numerical weather prediction, the long-term evolution of monthly foehn occurrence is still open to debate. Second, while many studies cited the importance of foehn winds on forest fires in single case studies, a ...Master Thesis -
Towards a Distributed Low‑Noise Signal Generation System for Quantum Gas Experiments
(2022)Master Thesis -
Bounded Cohomology of Groups acting on Cantor sets
(2022)We study the bounded cohomology of certain groups acting on the Cantor set. More specifically, we consider the full group of homeomorphisms of the Cantor set as well as Thompson's group $V$. We prove that both of these groups are boundedly acyclic, that is the bounded cohomology with trivial real coefficients vanishes in positive degrees. Combining this result with the already established $\mathbb{Z}$-acyclicity of Thompson's group $V$, ...Master Thesis -
V2X-Enabled Control of Connected Vehicles Over a 5G Network
(2022)Today’s research on automated driving systems shows a strong tendency to establish a more and more smartly connected environment. Communication between vehicles must be highly reliable and include low latencies to meet the stringent safety requirements of autonomous cars and the surroundings. The fifth generation (5G) networks will be the enabling technology to fulfill the challenging connectivity and serviceability demands of automated ...Master Thesis -
Optimal Stock Option Trading Strategies
(2022)Over the past decades, interest in finance, and more particularly interest in the evolution as well as the prediction of the markets, exploded. One aim that the automatic control laboratory at ETH Zürich would want to achieve is to use some of the robotic optimization models and to apply them to finance. During this internship, I only worked on the financial part of the problem. The main goal of my internship was to create some predictions ...Master Thesis -
-
In situ stress profiling along Bedretto boreholes via integrated geomechanical approaches
(2022)With the Bedretto Underground Laboratory for Geoscience, ETH Zurich has an excellent opportunity to gain a deeper insight into underground processes relevant for hydro-seismo-mechanical fractured granitic rock mass behaviour and subsequent applications. Improvement of state-of-the art equipment and research progress in geothermal energy, earthquake physics and rock stability can be achieved by conducting essential experiments directly ...Master Thesis -
A Distributed Hypergradient Algorithm for Intervention in Games
(2022)Hierarchical optimization problems, such as bilevel programs and Stackelberg games, are attracting increasing interest in both engineering and machine learning communities. Yet, existing solution methods lack either convergence guarantees or computational efficiency, due to the absence of smoothness and convexity in bilevel problems. We address this gap in the literature by designing a double-loop first-order method for single-leader ...Master Thesis -
Independent Mechanism Analysis in High-Dimensional Observation Spaces
(2022)Identifiability is a desirable property in the context of representation learning: in absence thereof, recovery of a ground-truth latent generative model based on observations is impossible. Identifiabilty has mostly been studied in the context of nonlinear Independent Component Analysis (ICA); in nonlinear ICA, identifiability of the generative model up to a meaningful equivalence class (e.g., permutation of latent factors) requires ...Master Thesis -
Numerical modelling of induced seismicity and faulting in visco-elasto-plastic rocks
(2020)Master Thesis -
Data-driven modeling of heat pumps and thermal storage units for MPC
(2022)The integration of renewable energy sources in the power grid poses new challenges for grid operators. Demand-side resources are promising candidates to address this challenge. Advanced control strategies can enable, such as peak shaving and load shifting. Heat pump systems coupled with thermal storage units acting as an energy buffer will play a vital role in supporting system operation in the future. As one example of advanced control ...Master Thesis -
Learning and Control for Exoskeletons
(2022)Upper limb exoskeletons are assistive robots used to support subjects with physical disabilities. The key element of these exoskeletons is a controller which ideally provides required portion of the torque in the intended direction, and ensures the synchronization of the movement between the robot and the user. While achieving these goals, it is necessary that the robot is controlled in a smooth and human-like manner. In this thesis, we ...Master Thesis -
Scalable adaptive model predictive control with parameterized tubes
(2022)This work considers linear systems with uncertain parameters and additive disturbance. Two frameworks for the adaptive model predictive controller (MPC) with different tube parameterizations are introduced to guarantee robustness for uncertain parameters and disturbances. The set membership identification approach, which uses a membership set to bound the uncertain parameter set, is applied to reduce the uncertainty. Moreover, a method ...Master Thesis -
Iterative learning control for uncertain systems satisfying Lipschitz inequalities
(2022)Iterative learning control, or ILC for short, is a feedforward control technique that aims at iteratively improving the performance of a system that operates repetitively by learning from previous trials. Recently, attention has been focused on optimization based ILC, where input and state constraints can be considered explicity. The literature, however, seems to lack ILC schemes that are able to guarantee constraint satisfaction in the ...Master Thesis -
Distributed MPC for Demand Side Management using Smart Homes
(2022)Building comfort control contributes a substantial quantity of greenhouse gas emissions. Two approaches can be combined to decrease this carbon footprint using model predictive control (MPC) strategies. First, the energy consumption of already-installed convectors should be reduced while maintaining thermal comfort. Secondly, it may be advantageous for the distribution grid to incorporate more renewable energy sources and to limit the ...Master Thesis -
Enforcing Fairness in Autonomous Peer-to-Peer Energy Hub Trading
(2022)Energy hubs are multi-carrier energy systems that incorporate generation, conversion, and storage technologies. They are gaining increasing relevance in the energy sector due to their ability to match local production with consumption, incorporate renewable energy sources and increase the flexibility of supply. We investigate autonomous peer-to-peer trading between energy hubs to further optimize the energy usage within the network, while ...Master Thesis -
Identification and MPC control of a hydrogen energy storage system for peak shaving EV demand
(2022)The penetration of electrical vehicles (EVs) in the past years has raised some concerns about the response of the electrical grid to the sensitive increase in demand they brought. In particular, spikes of power during time windows could lead to instability issues in the distribution network. Additionally, high power peaks are charged with high electricity costs. This project addresses the problem of minimizing the daily power peak of an ...Master Thesis