Physics

Title of thesis:&nbsp;<span style="font-size: 12.0pt; font-family: 'Aptos',sans-serif; mso-fareast-font-family: Calibri; mso-fareast-theme-font: minor-latin; mso-bidi-font-family: Aptos; color: black; mso-ansi-language: SV; mso-fareast-language: SV; mso-bidi-language: AR-SA;">Voltage prediction and uncertainty estimation using machine learning on fuel cells
<span style="font-size: 12.0pt; font-family: 'Aptos',sans-serif; mso-fareast-font-family: Calibri; mso-fareast-theme-font: minor-latin; mso-bidi-font-family: Aptos; color: black; mso-ansi-language: SV; mso-fareast-language: SV; mso-bidi-language: AR-SA;">Abstract: 
With an increasing need for electrification in the transportation sector, fuel cells are gaining prominence as a clean energy solution. Fuel cells produce electricity from hydrogen and oxygen, with the only by-product being water and heat. One step to further improve the technology is to model the fuel cell stack. The aim of this thesis is to investigate how machine learning can be used to predict the output voltage of a fuel cell stack, while also providing an uncertainty estimation of the prediction. This was done by evaluating four machine learning models, specifically: a Gaussian Process Regressor (GPR), Monte-Carlo Dropout (MCD), a Deep Ensemble (DE) and finally a Functional Nearest Neighbor - Hypernet Variational Inference (FuNN-HyVI) model. These models were compared in terms of accuracy, out-of-distribution classification and residual-uncertainty correlation.&nbsp; &nbsp; All models were found to accurately predict the output voltage, with the best model being the DE. The GPR excelled at out-of-distribution classification, with the DE and FuNN-HyVI models also showing good results. The results of residual-uncertainty correlation are inconclusive but indicate that all models have potential to perform well with some further modifications. Overall, the GPR emerged as the most balanced model across all metrics and, due to its simplicity and ease of implementation, stands out as the most practical choice for real-world applications.&nbsp;
Examiner: Mohsen Mirkhalaf Supervisor: Stefan Bohatsch Opponent: Mathias Arvidsson
<span style="font-size: 12.0pt; font-family: 'Aptos',sans-serif; mso-fareast-font-family: Calibri; mso-fareast-theme-font: minor-latin; mso-bidi-font-family: Aptos; color: black; mso-ansi-language: SV; mso-fareast-language: SV; mso-bidi-language: AR-SA;">Location of presentation: <a rel="noopener" href="https://maps.chalmers.se/#495a8bdb-ae0d-4c94-8be5-68240f9a5686" target="_blank" title="Chalmers maps" data-anchor="#495a8bdb-ae0d-4c94-8be5-68240f9a5686">von Bahr, Soliden floor 3</a>

Title of thesis: Designing a Modular Binding System for Alpine Skiing
Abstract:
This master’s thesis focuses on developing a digital function model for a modular alpine binding system using Catia software. The primary objective is to design a versatile binding system that addresses the diverse needs of skiers in both resort and off-piste environments. Current market offerings often involve compromises: downhill ski bindings fail to meet the demands of ski touring, and vice versa. This highlights the growing demand for a modular binding system that provides optimal performance for different ski styles and in different environments by enabling the exchange of specific binding components. The research includes an analysis of existing products, competitor offerings, and user needs to define key design requirements. Methods such as functional analysis and concept generation have been employed to conceptualize the system in Catia. Particular emphasis is placed on identifying high-value features for customers, understanding usage environments, and ensuring modularity without compromising performance. Preliminary concepts demonstrate the potential for a modular binding system to enhance performance regardless of skiing style or environment. However, due to the complexity of the system and its components, further research and an extended timeline are necessary to fully develop and validate the functionality of all exchangeable parts.
Examiner: Magnus Karlsteen Supervisor: Magnus Karlsteen Opponents: Carl Ingemarsson and Eric Fabricius
Location of presentation: <a rel="noopener" href="https://maps.chalmers.se/#24d8b7be-1f59-4388-be28-a07861b429b2" target="_blank" title="Chalmers Maps" data-anchor="#24d8b7be-1f59-4388-be28-a07861b429b2">PJ-salen</a>

Architecture and Civil Engineering

Bachelor’s thesis Chemistry and Chemical Engineering

Chemistry and Chemical Engineering

Communication and Learning in Science

Computer Science and Engineering

Electrical Engineering

Industrial and Materials Science

Life Sciences

Mathematical Sciences

Mechanics and Maritime Sciences

Microtechnology and Nanoscience

Space, Earth and Environment

Technology Management and Economics

Presentation of master thesis

Choose subject

Course syllabus for

Subject

Each academic year contains different courses

See programme page at universityadmissions.se

Educational area

Permissible substitutions of compulsory courses

To see your programme syllabus – select the academic year in which you started the programme.

[en: page.studyprogrammepage.syllabusesview.coursesdescription]

Presentation of master thesis

Master thesis presentations at Physics

January 13 2025

16:00- 17:00 August Kälvesten

January 27 2025

13:00- 14:00 Johanna Steninger, MPPDE