Chalmers Open Digital Repository

Välkommen till Chalmers öppna digitala arkiv!

Här hittar du:

Studentarbeten utgivna på lärosätet, såväl kandidatarbeten som examensarbeten på grund- och masternivå
Digitala specialsamlingar, som t ex Chalmers modellkammare
Utvalda projektrapporter

Forskningspublikationer, rapporter och avhandlingar hittar du i research.chalmers.se

Enheter i Chalmers ODR

Välj en enhet för att se alla samlingar.

Visar 1 - 2 av 2

Senast publicerade

Autonomous Navigation In Real-Time Endovascular Simulation
(2026) Magnusson, Niklas; Pettersson, Lukas
This Master’s thesis investigates the use of deep reinforcement learning for autonomous navigation in a simulated endovascular environment. Specifically, a Soft Actor-Critic (SAC) algorithm is employed to train an agent to control a micro guidewire and a micro catheter for path-following tasks. The simulation environment is based on the VIST simulator, where the agent observes a 23-dimensional state representation capturing relevant path-following information. The agent was rewarded for path following, with the reward function incorporating vessel centerline alignment, penalization of deviation from the path, and progression toward the target. The agent produces continuous translation and rotation commands for the respective tools. Training was conducted on four anatomies over 1600 episodes, each consisting of up to 500 time steps depending on task completion. The best performing model emerged after 800 training episodes, which achieved a success rate of 94 % on validation data. During testing, it achieved a 96 % success rate on an unseen cerebral anatomy and an 88 % success rate on an unseen liver anatomy. This indicates that the agent learned transferable navigation strategies rather than anatomy specific memorization, which demonstrates that deep reinforcement learning is a viable approach for endovascular navigation in a simulated environment.
Towards Reliable Retrieval Systems: Design and Evaluation of an Agentic GraphRAG Pipeline
(2026) Spreitz, Adam
Large language models are increasingly used for knowledge-intensive question answering, but their reliability remains limited when answers must be grounded in domain-specific documents. This limitation is particularly important in high-assurance environments, where systems must be controllable, traceable, and deployable without relying on external infrastructure. This thesis investigates the design and evaluation of an agentic GraphRAG system for document-grounded question answering, using scientific literature as a controlled proxy for technical internal documentation. The implemented system combines document ingestion, section-aware chunking, knowledge graph construction, vector indexing, graph traversal, reranking, and languagemodel- based answer generation. Two retrieval architectures are compared under shared conditions: a VectorRAG baseline using iterative hybrid search and a GraphRAG system using community-first hierarchical traversal over an explicit knowledge graph. The systems are evaluated across multiple retrieval configurations, generation models, and query sets using automated RAG evaluation metrics, statistical tests, and pairwise LLM-as-judge comparisons. The results show a clear divergence between retrieval-oriented metrics and answerlevel evaluation. VectorRAG achieves stronger automated retrieval scores, particularly on context precision and recall, while GraphRAG is preferred in holistic answer comparisons and manual validation. This suggests that chunk-level retrieval metrics do not always capture the usefulness of structurally retrieved evidence for downstream answer generation. The findings indicate that graph-based retrieval can improve answer quality and interpretability in agentic RAG systems, but also highlight important limitations related to dataset construction, evaluator dependence, corpus scale, graph quality, and agentic control.
Konceptutveckling av ett belastningsgränssnitt - En användarcentrerad produktutvecklingsprocess för en optimerad squat
(2026) Hansson, Alicia; Sandell, Linnéa
This report documents the development of a load interface for belt-squat machines, conducted on behalf of a fitness equipment company. The purpose of the project was to investigate and develop a concept that maximizes comfort and usability for a broad target group for usage in commercial gym environments. Through a user-centered design process, including market analyses and user testing of existing products, the areas for pain at pressure points, instability, and a lack of user-friendliness were identified as objects for development. The work resulted in three conceptual solutions that were evaluated against a requirement specification, where the V-band concept was selected as the final design proposal. The result is a detailed-designed belt-squat belt that implements a user-friendly interface design. Through a combination of a stable backplate and a unique, foldable geometry, the pressure distribution over the hips and lower back is optimized without restricting the user's range of motion. To lower the barrier to use, the belt features a numbered adjustment system for specific settings and a permanent anchoring to the machine, which minimizes the risk of misuse and misplaced equipment. The concept has been dimensioned based on anthropometric data to include users within the 5th to 95th percentiles of the Swedish population. The conclusion of the project is that by prioritizing user comfort, mechanical simplicity, and an intuitive visual interface, the user experience of a belt-squat belt can be significantly improved, thereby making the execution of the exercise more accessible and safe for a universal target group.
Design and Evaluation of Efficient Power Amplifier Architectures for Millimeter-Wave 5G Backhaul Applications
(2026) Shinde, Dhanashree Ashok
The rapid global rollout of 5G mobile communications has put a great strain on millimeter-wave backhaul infrastructure, requiring power amplifiers (PAs) with high output power, high efficiency, and wide bandwidth. This thesis presents the design and simulation evaluation of an efficient Class AB power amplifier for 5G point-topoint backhaul links in the 37–40 GHz band. The design is implemented using the United Monolithic Semiconductors GH10-10 Gallium Nitride (GaN) High Electron Mobility Transistor (HEMT) process design kit in the Keysight Advanced Design System simulation environment. The methodology makes use of a two-stage cascaded architecture with a driver stage and a four-transistor parallel power amplifier stage. To bridge the gap between the theoretical performance and physical realization, the design incorporates the electromagnetic (EM) model-based layouts for the input, interstage, and output matching networks. The integrated RC stabilization networks were also optimized to suppress low-frequency and out-of-band instabilities, which are common in high-gain GaN devices. The final EM-realized MMIC simulation results show a saturated output power of 34.6dBm (2.9W) at 37 GHz and 33.15dBm at 40GHz. The design exhibited a peak Power Added efficiency (PAE) of 25.93% and 10 dB output power backoff PAE of ∼ 5%. The small-signal performance shows gain (S21) ranges between 11.3 dB and 12.1 dB over the bandwidth and an input return loss (S11) better than 13.6 dB. The final MMIC is found conditionally stable from 1 to 50 GHz with Rollett stability factor also using advanced stability verification with Ohtomo loop gains and Kurokawa driving-point admittance. The results validate the effectiveness of GaN HEMT technology to construct robust and energy efficient transmitter front-ends for next-generation 5G backhaul infrastructure. These results are lower than the design goal and can be improved in future work by adding an additional driver stage or by increasing the total gate periphery of the driver-stage transistors to provide a higher drive capability. Such changes would require re-design of the matching networks to compensate for increased device parasitics.
Estimation of Iron and Zinc Bioavailability in Salad Formulations with Different Protein Sources
(2026) Lagerroth, Elin
A plant-based dietary shift has the potential to reduce the climate impact of the food systems. However, this transition must not compromise the uptake of essential nutrients. Salad bars play a role in this transition, offering customizable and plant rich convenience meals. Despite their popularity, there is limited knowledge regarding the bioavailability of essential minerals in these composed salad meals. This study estimated the bioavailability of iron and zinc in three salad compositions containing different protein sources (meat/fish, lacto-ovo, and plant-based) to determine if they meet the nutritional requirements of school lunches for teenage girls, who are highly susceptible to mineral deficiencies. The salad components were analyzed for phytic acid, iron, and zinc content using high-performance ion chromatography (HPIC) and atomic absorption spectroscopy (AAS). Mineral bioavailability was subsequently estimated using established algorithms (Hallberg and Hulthén, Miller) and molar ratios. The results demonstrate that while the meat-based composition provides highly absorbable iron and zinc, the plant-based and lacto-ovo alternatives fail to meet the absorbable iron requirements for individuals with high physiological needs. This is potentially due to high concentrations of inhibitors such as phytic acid and calcium. Conversely, all salad compositions had high absorption efficiency of zinc. These findings suggest that further optimization is required for salad-based meals to be nutritionally adequate with regard to iron. Potential strategies include substituting meat analogs with fermented products like tempeh, which has a reduced phytic acid content, or utilizing non-soy-based alternatives. Furthermore, mineral bioavailability can be enhanced by moderating the intake of phytate-, calcium-, soy-, and egg-rich components, while increasing the presence of enhancers such as ascorbic acid or animal proteins.