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 inlagda

Environmental Perception for Autonomous Forestry Vehicles

(2025) Olsson, Viktor; Skoog, Hannes

Abstract Autonomous navigation in forestry environments presents significant challenges due to complex, unstructured terrain with varying visibility conditions. This thesis presents a novel sensor fusion approach integrating LiDAR and stereo camera data for enhanced terrain mapping in forestry applications. The project develops an uncertainty-aware fusion framework based on Kalman filtering that effectively combines the high accuracy of LiDAR with the dense coverage of stereo camera data, while properly accounting for each sensor’s unique error characteristics and uncertainties. Additionally, a dynamic voxel-based representation is implemented that adapts map resolution to terrain complexity, optimizing memory usage while maintaining high fidelity in regions of interest. Experimental results demonstrate measurable improvements in various dimensions: the dynamic voxelization reduced memory usage by 31.65% and improved map update time by 44.27% compared to traditional fixed-size voxel grids, while maintaining mapping quality. Testing on real-world autonomous navigation routes showed that the proposed approach enables more complete trajectory following compared to the previous single-sensor approach, achieving path lengths significantly closer to the planned trajectory - for instance, 38.99m compared to 18.99m in one test. This work demonstrates that intelligent fusion of complementary sensors, combined with adaptive mapping techniques, can significantly improve terrain perception for autonomous vehicles operating in challenging off-road environments.

Impact of Dynamic Gate Drive on System Level Efficiency of SiC-Based ElectricDrive Units

(2025) Murnieks Andersson , Sara

Abstract The Electric Drive Unit (EDU) is a critical subsystem in Battery Electric Vehicles (BEVs), responsible for converting electrical energy from the battery into mechanical power for propulsion. A central element of the EDU is the inverter, which converts DC power into AC to drive the electric motor. Silicon Carbide (SiC) inverters are increasingly employed in BEVs due to their high efficiency and thermal performance. However, their fast switching characteristics introduce challenges such as voltage overshoots, current spikes, and Electromagnetic Interference (EMI), potentially affecting system reliability and efficiency. This thesis investigates a Dynamic Gate Drive (DGD) strategy, in which the external gate resistance is adjusted for the turn-off transition of SiC MOSFETs to balance switching losses and voltage overshoot. Circuit-level simulations were performed using a Double Pulse Test (DPT) setup in LTspice to evaluate switching behaviour, while conduction losses were estimated based on the relationship between device voltage rating and on-state resistance RDS. Motor performance under different operating conditions was assessed in Motor-CAD, and system-level losses was analysed under a Worldwide Harmonized Light Vehicle Test Procedure (WLTP) drive cycle. The results indicate that the DGD approach can reduce inverter losses, particularly under low initial state-of-charge (SOC) conditions, and provide marginal improvements in overall EDU efficiency. However, the impact of DGD on the efficiency of the electric machine was not straightforward, and no clear trend could be established. Nevertheless, reduced voltage overshoot allows for a modest increase in DC link voltage utilization, contributing to higher base speed, increased torque output, and improved peak power performance in the constant power region. While the improvements at the system level are limited, the findings highlight that, in principle, DGD is a method to enhance inverter performance and enable more effective utilization of existing powertrain components in electric vehicle applications.

Electric machine loss measurement and modeling - PWM versus sinusoidal feeding

(2025) Moänge, Arvid; Liseau, Philip

Abstract As the electrification of the automotive industry continues, the importance of high efficiency electric machines increases significantly. The ability to accurately predict the efficiency and losses in the designing stage is therefore essential. This thesis studied the degradation in magnetic properties in silicon steel strips cut with different widths. These pieces were then used to develop a model to implement manufacturing effects due to cutting in to the electric machine simulations to more accurately model the iron losses. The simulations were fed with both a sinusoidal current and a pulse width modulated voltage feeding to compare the effect of the different feeding types on top of the cutting degradation effect, and compared with each other. For all points simulated, the core losses increased when a layer of degraded material closest to the cut edges of the stator was used, up to 25% with the sinusoidal feeding. For the pulse width modulation case, five different operating points were simulated, showing an increase in core losses of up to 7% when using degraded material. Comparing the effect the feeding types had versus each other, the pulse width modulation fed simulations had core losses up to 258% higher than the sinusoidally fed case. One measurement of the actual machine was compared with the pulse width modulated feeding simulations where the difference between losses was 6.7%, showing promise in the model for the simulated points.

Sequential Graph-Based Decoding of the Surface Code using a Hybrid Graph and Recurrent Neural Network Model

Fjeldså, Ole Aleksander Larsen; Jonasson Johansson, Gustaf

In order to achieve reliable quantum computation with noisy qubits, quantum error correction (QEC) is necessary. Quantum error correcting codes mitigate the inher ent noise in quantum systems by distributing the logical state over several qubits, thereby introducing redundancy. One such promising code is the surface code. It encodes the logical qubit using a two-dimensional lattice of physical data qubits and ancilla qubits. By taking and decoding measurements on the ancilla qubits of the surface code, one can deduce whether a logical bit- or phase-flip has occurred. However, this is a complex and potentially time-consuming task. Multiple decoding algorithms exist, such as the classical minimum-weight perfect matching (MWPM) decoder. In recent years, data-driven algorithms have been shown to decode the surface code with a high degree of accuracy. In this thesis, we present a machine learning approach to decoding the surface code using a combination of graph neural networks (GNN) and recurrent neural networks (RNN). Specifically, graph represen tations are constructed over a short, sliding time window of syndrome measurement data. Each representation is processed by a GNN and its output is used as a learned high-dimensional embedding for an RNN. This enables continuous decoding of mea surement patterns over longer time series. While the decoder is trained on relatively short syndromes, it is able to generalize for unseen syndromes and longer time se ries, outperforming the classical MWPM algorithm across both short and long time series. This work opens up a new approach to reliable and potentially fast decoding of QEC codes.

Applicerbarhet av Lean i en butiksmiljö

(2025) Voigt, Lucas; Holmén, Olle

Under de senaste decennierna har Lean etablerats som en filosofi som utmärker sig för förbättrade flöden, standardiserade arbetssätt, skickliga mänskliga resurser och ständiga förbättringar inom tillverkningsindustrin. Studien syftar till att undersöka hur detta kan appliceras i en butiksmiljö, med särskilt fokus på ICA Malmborgs Mobilias kökslinje. Genom en fallstudie med observationer, intervjuer värdeflödeskartläggningar och spaghettidiagram analyseras flöden, standardiserat arbetssätt, organisering av mänskliga resurser och förbättringsarbete ut ett Lean perspektiv. Analys av kvalitativa och kvantitativa data visar att förbättringspotentialen är betydande inom vardera av de undersökta områdena. Onödiga rörelser, väntetider, avbrott och bristande struktur kännetecknar flöden. Det förekommer standardisering i begränsad utsträckning, men där instruktioner sällan används i praktiken. Standardisering av vanor, rutiner och tankesätt är dess värre. Personalens engagemang hämmas av otydligt ledarskap, låg återkoppling och avsaknad av struktur för delaktighet. Det förbättringsarbetet saknar både ansvarsfördelning och uppföljning samt domineras av kortsiktiga finansiella mål snarare än långsiktigt värdeskapande. Trots dessa utmaningar visar studien att Lean är applicerbart även i en butiksmiljö, under förutsättning att principerna anpassas till verksamhetens unika förutsättningar. Genom att införa daglig styrning, tydlig standardisering, närvarande ledarskap och strukturerade förbättrings vanor kan ICA Malmborgs Mobilia skapa ett mer stabilt, engagerat och värdeskapande arbetssätt. Studien bidrar även till det akademiska fältet genom att fördjupa förståelsen för Lean i detaljhandeln, ett område där tidigare forskning varit begränsad.