Chalmers Open Digital Repository
Not a member yet
26247 research outputs found
Sort by
Optimal planning of data centers with on-site generation and storage a case study in Dublin Ireland
No full textData center energy demand is soaring globally. In Ireland, data centers accounted
for 21 percent of national electricity demand in 2023 and are expected to represent
31 percent by 2030, with virtually all incremental load in Dublin. These pressures
have driven the Commission for Regulation of Utilities (CRU) to require on-site
dispatchable generation (and/or storage) for all pending data center approvals and
led EirGrid, the Transmission System Operator (TSO) in Ireland, to suspend new
data center applications until 2028. To address these challenges, this thesis develops
a mixed-integer linear programming model from first principles to size and operate
an on-site energy portfolio consisting of photovoltaic panels, onshore wind turbines,
small modular reactors, and battery energy storage for a 5 MW data center in
Dublin, Ireland. The model minimizes annualized life-cycle cost by co-optimizing
capacity investments and hourly dispatch under realistic time-of-use tariffs, whole sale
spot prices, load profile, operational constraints, and regulatory requirements.
Under 2025 cost assumptions, the cost-optimal mix comprises PV and wind with
grid imports. Battery energy storage enters the least-cost portfolio by 2028 on pure
energy arbitrage. Including additional revenue streams, such as demand response,
would enable BESS deployment as early as 2025. Sensitivity analyses reveal that
system scale, resource cost trajectories, spot price volatility, and demand response
participation can substantially reshape investment decisions: SMRs become competitive
in the 50 MW scenario; wider intraday price spreads alone justify significant
storage capacity; and dynamic demand response revenues can more than double
BESS earnings compared to arbitrage. These results demonstrate the model’s utility
as a decision-support tool for data center developers, investors, and planners
navigating complex economic, technological, and regulatory uncertainties
Quick Access - Optimizing AOT Compiled Dynamic Programming Languages
No full textResearch on dynamic programming languages typically focus on just-in-time (JIT) compilation. However, JIT techniques are not always viable due to hardware limitations or security concerns, making it valuable to study ahead-of-time (AOT) compilation of dynamic languages as well. The GameMaker game engine targets mobile devices, game consoles, and WebAssembly — all platforms poorly suited for JIT compilation. Instead, GameMaker supports AOT compilation of JavaScript and its own language, GameMaker Language. In this thesis, we extend the engine with new profiling tools to analyze the memory usage and runtime performance of property accesses, a frequent and performance-critical operation in dynamic languages. Using our new tools, we identify several opportunities for improvement in both the compiler and runtime environment. Finally, we propose and implement solutions to the identified areas of improvement. These include polymorphic property caches, a pipeline for profile-guided optimizations, caching of accessor properties, and cache pools designed to facilitate cache invalidation. In certain benchmarks, our solutions achieve speedups with factors ranging from 1.5 to 2.9
En analys av närsjukvårdens arbete i Göteborg
No full textThe healthcare system faces significant challenges, with a pressing shortage of hospital beds
and a growing need for healthcare services. One potential solution to face this challenge is
home care, which can reduce the burden on existing healthcare facilities and improve
resource efficiency. By treating patients pre-hospitally and giving those possible treatments at
home, the risk of them requiring hospital admission is reduced, thereby also reducing
pressure on healthcare staff. In this context, mobile healthcare teams play a crucial role in
delivering this care at home.
This project aims to analyze the work of mobile healthcare teams at Östra Hospital, Mölndal
Hospital, and Sahlgrenska Hospital within the Sahlgrenska University Hospital network in
Gothenburg. With a particular focus on understanding the shared practices and collaboration
between the hospitals, as well as identifying barriers and opportunities for improvement. The
project was carried out through a combination of qualitative research methods including
semi-structured interviews and observations with staff working within the mobile teams and a
literature review.
The findings indicate that while each hospital has its own strengths and practices, differences
in resources and prior experience have led to distinct working methods. The differences in
resources means staff, equipment and technologies. While all teams share the overarching
goal of providing equitable and high-quality care to all patients in Gothenburg, varying
approaches have led to friction between them, particularly in how they prioritize urgent
versus planned care and handle communication. For example, Östra Hospital is more focused
on managing urgent visits and has greater resources, while Mölndal emphasizes long-term
care, particularly for elderly patients. Sahlgrenska strikes a balance between urgent and
planned care. Despite these differences, all teams are highly competent and dedicated to their
work, which ultimately contributes to the overall success of the home care initiative.
The study underscores the need for a shared vision and improved coordination among the
teams. Collaboration, common guidelines, and shared goals are essential to overcoming
friction and ensuring long-term success and organizational learning. Recommendations
include regular workshops, feedback systems, and performance metrics to foster learning and
knowledge-sharing. By leveraging each team’s strengths and optimizing resource allocation,
the hospitals can establish a cohesive approach to mobile home care, which will play an
increasingly vital role in healthcare’s future
Strategiska vägval för västsvensk petroleumraffinering: En studie av hur utsläppens karaktär utgör en drivkraft för grön omställning
No full textProblem
Allt större klimatförändringar innebär förändrade konsumtionsmönster för fossila bränslen, till
följd av strängare regulatoriska krav och en växande efterfrågan på förnybara energikällor.
Detta hotar petroleumraffineringsindustrins framtida existensberättigande, trots att branschens
produkter i många fall är icke-substituerbara. Om framtida efterfrågan ska tillgodoses hållbart,
måste petroleumraffinaderier förändra sina strategier för att förbli långsiktigt och
klimatmässigt konkurrenskraftiga.
Syfte
Syftet med studien är att förklara och analysera hur västsvenska petroleumraffinaderier
anpassar sina strategier för att förbli konkurrenskraftiga, samtidigt som de påverkas av den
gröna omställningen och föränderliga marknadsförhållanden. Studien syftar till att kartlägga
de strategiska åtgärder och investeringar som vidtas för att anpassa petroleumraffinaderiers
kärnverksamhet när deras fossilbaserade produkter riskerar att bli förlegade och irrelevanta.
Teoretiskt ramverk
Studiens huvudsakliga forskningsområden är affärsutveckling och strategi samt institutionell
teori. Inom dessa för studien relevanta forskningsområden utgörs det teoretiska ramverket av
vetenskapliga artiklar, böcker och branschspecifika publikationer. Källornas tillförlitlighet
bedömdes med hänsyn till de fyra källkritiska kriterierna och studien har premierat artiklar
publicerade av välrenommerade och tillförlitliga universitet, tidskrifter och myndigheter.
Metod
Fallstudien var utforskande till sin natur och avgränsades till att undersöka fyra västsvenska
petroleumraffinaderier, verksamma inom samma huvudsakliga bransch men inom olika
delbranscher. Primärdata från de undersökta företagen samlades in genom intervjuer med
anställda med strategisk insikt och kompletterades med sekundärdata från främst företagens
års- och hållbarhetsredovisningar. Teoretiska modeller, analysverktyg och tidigare forskning
användes för att analysera insamlade data samt för att underbygga diskussion och slutsatser.
Resultat och implikationer
Studien bidrar teoretiskt till forskningsområdena affärsutveckling och strategi samt
institutionell teori, med förståelse för hur västsvenska petroleumraffinaderier agerar strategiskt
när de hotas av externa faktorer. Studiens praktiska bidrag utgörs av en kartläggning av
hållbarhetsinriktade strategiförändringar som är relevanta för aktörer och regulatorer inom
petroleumraffineringsindustrin under en grön omställning. Drivmedelsproducenternas
strategiförändring utgörs främst av investeringar för en diversifiering av produktportföljen med
förnybara insatsvaror som reducerar konsumtionsrelaterade utsläpp. Producenterna av
specialiserade petrokemiska produkter gör endast marginella investeringar i förnybara
alternativ, till förmån för att energieffektivisera produktionsprocessen och för att skapa
cirkulära flöden för fossila produkter. Konsumtionsrelaterade utsläpp resulterar i faktorer som
innebär att drivmedelsproducenterna genomför mer omfattande strategiska förändringar än
producenterna av specialiserade petrokemiska produkter
Modeling Protein-Ligand Binding Affinity Using Graph Neural Networks: Integrating Molecular Interactions and Physics-Based Properties
No full textPredicting protein-ligand binding affinity has always been one of the primary challenges in drug discovery. Though many machine learning approaches have been applied and reached substantial progress, despite advancements, the accurate prediction by integrating molecular interactions and physics-based properties of ligands and proteins remains challenging. In this project, we present a graph neural networks (GNNs)-based framework for predicting protein–ligand binding affinity, using publicly accessible CrossDocked2020 dataset. Our project compares three message-passing architectures—Linear, Set Transformer Aggregation (STA), and Graph Attention Network (GAT). Our best model achieves performance of Pearson’s R ≈ 0.79 and Kendall’s τ ≈ 0.58. We present a practical GNNs-based framework with plausible binding affinity prediction capabilities, designed to effectively differentiate correct poses from incorrect ones
Preserving Semantics of Multi-Threaded Programs During Cross-ISA Dynamic Binary Translation
No full textDynamic Binary Translation (DBT) is a method used to emulate programs on platforms on which they cannot execute natively. In the past, DBTs either did not emulate multi-core programs or did not parallelize their execution. This is no longer the case, as modern processors are often multi-core, necessitating better scaling in DBTs. Renode [1] is one such DBT that is able to emulate multi-core programs using parallel execution. However, Renode — like many other DBTs — fails to correctly emulate the semantics of certain atomic instructions. In particular, emulation of the RISC-V instructions Load-Reserved (LR) and Store-Conditional (SC) is currently incorrect. These semantics are paramount for program correctness.
In this thesis, we improve Renode’s correctness by applying the Hash table-based Store-Test (HST) — a scheme proposed by Zhao et al. [2] — to correctly emulate LR/SC instructions. Using model checking, we find that implementing HST as described by Zhao et al. in Renode results in a race condition. We show how to remediate this race condition in Renode.
Furthermore, we compare the performance of two HST implementations: one written directly in an intermediate representation (IR) similar to assembly, the other written in C using helper functions. Previous work suggests that IR is faster due to less runtime overhead, which we show holds in this case. We find that the IR implementation is 34% faster than helpers in microbenchmarks and 6–18% faster in the PARSEC [3] benchmark suite.
Our IR implementation of HST in Renode improves both correctness and scalability. We show that our implementation can boot Linux on an embedded platform with multi-core emulation enabled, which Renode in its current state (current Renode) cannot do due to correctness issues. Moreover, our implementation scales well when current Renode does not: in an 8-thread microbenchmark of LR/SC, our implementation is 15.6x faster than current Renode. We find that this scalability can be achieved with as little as 8 KiB of extra memory usage
Strömningssimulering av fendrift: Fluid-struktur-växelverkan med deformerbar fena
No full textDetta kandidatarbete syftar till att undersöka och modellera fendrift, ett biomimetiskt
alternativ till traditionella propellrar, genom utveckling av en beräkningseffektiv
CFD-simuleringsmodell. Arbetet fokuserar på att efterlikna rörelsemönstret hos
en delfinfena, med hjälp av simuleringsverktyget WaterLily och en egenutvecklad
FEM-lösare i Julia. Målet var att skapa en parametriserbar modell som kan användas
för att analysera och optimera fenans design och rörelse, samt att utvärdera
WaterLilys lämplighet som undervisningsverktyg i strömningsmekanik. Projektet
resulterade i en simuleringsmodell som möjliggör visualisering av fenans rörelse och
vissa grundläggande kraftanalyser, men visade också på begränsningar i både modellens
noggrannhet och simuleringsverktygets kapacitet. Slutsatsen är att fendrift
har potential som ett mer miljövänligt och effektivt framdrivningssätt, men att mer
avancerade och precisa simuleringar krävs för tillförlitliga resultat
An Interactive Map of the World’s Languages
No full textThis project thesis presents the development of an interactive language map that
visualizes the global distribution of spoken languages. The system integrates
linguistic data from Wikidata and geospatial data from OpenStreetMap to allow
users to search for languages and view where they are spoken. The application was
built with a React and TypeScript frontend, a NestJS backend, and a MongoDB
database hosted in the MongoDB Atlas cloud. SPARQL queries were used to extract
structured language data, which was cleaned and stored to enable fast queries.
The resulting product enables interactive visualization of language regions with
support for filtering by country or region, viewing speaker statistics, and language
family information. The map also supports interactive exploration by clicking on a
country to display information about the languages spoken in that country. The
project demonstrates how open data sources can be used to create educational
visualizations, although the quality and coverge of the external databases constrain
the final accuracy
The swedish transport administration’s climate requirements in the procurement of infrastructure and bridge projects
No full textIn line with increasing climate ambitions, the Swedish Transport Administration (STA)
has introduced stricter climate requirements in its procurement of infrastructure and
bridge projects since 2015. Requirements for climate reductions are gradually
increasing until the final goal of climate neutrality is achieved in 2040, which is
pressuring contractors and consultants to change their working methods, choice of
materials, and technical solutions. This thesis investigates how STA’s climate
requirements are currently perceived and managed by contractors and consultants, in
order to explore potential improvement measures in setting climate requirements in
public procurement of infrastructure and bridge projects. There is limited previous
research in this area, which makes this thesis a valuable contribution to the ongoing
transition for industry stakeholders. The aim of the study is to provide valuable
information to the client, the Swedish Transport Administration, in order to gain a better
understanding of how contractors and consultants experience their climate requirements
and thereby identify obstacles to improve their formulation of climate requirements. An
abductive research method was chosen, where an interview study was conducted in
combination with the development of a theoretical framework. A total of twelve semi structured interviews were conducted with various actors active in the construction
sector. Two infrastructure and bridge projects were selected as contextual examples,
where one representative from each party – client, contractor, and consultant – was
interviewed. The remaining six interviewees were selected based on their extensive
experience and expertise in the construction sector, with a particular focus on issues
related to climate requirements in public procurement. The results show that climate
requirements which are material- or fuel-specific are perceived as clear and feasible,
whereas the percentage-based reduction requirements are considered complex to
implement and difficult to follow up. Consultants request guidance and more
standardized calculation tools, while contractors call for clearer incentives that better
reflect the additional costs imposed by the climate requirements. The conclusion
emphasizes the importance of involving climate aspects early in the projects, increased
collaboration and knowledge exchange between actors, and more flexible and
innovation-promoting climate requirements. The study provides key recommendations
for improving the effectiveness of climate requirements to reduce emissions and
contribute to a sustainable construction sector
Controlled multi-body dynamic simulation for structural characterization
No full textVirtual simulation is a popular tool used in the design of modern drivetrains, potentially offering a more efficient iterative design approach than hardware testing. For example, using feedback controls in mechanical system simulations allows testing alternative drivertrain models with different control strategies at a minimal cost. However, feedback control functionalities are seldom included in simulations for mechanical
system design analysis and vice versa. Finding efficient ways that bridge the gap between the fields of structural design and feedback control design is of central interest for this project.
Phase Locked Loop (PLL) and Control Based Continuation (CBC) are feedback control architectures that recently gained popularity within the structural analysis community for characterizing nonlinear behavior. The basic analysis includes mapping of Frequency Response Curves (FRC) and the Backbone curve of the chosen dynamical system. PLL and CBC have been tested by the research community and are known to work, in both virtual and physical environments, for systems with relatively few Degrees Of Freedom (DOF). DOF can be defined as the minimum number of states required to model the system. A typical drivetrain model of today can have on the order of thousands of DOFs, depending on analysis purpose and chosen level of fidelity (higher fidelity models are typically more accurate but also use a larger number of DOFs). So, to deliver efficient feedback control of larger structural models, there is much scope for improvements among existing control architectures to satisfy industry requirements.
The thesis work aims to re-implement existing control architectures in a simple 1- DOF Duffing oscillator model. Another goal is to check whether the model-free controller used in the CBC control architecture can be replaced by a model-based controller. Model-based controllers use predicted system dynamics to generate optimal actuator signals. They provide better results, but are generally harder to implement. Finally, the process of integrating newly developed controllers into the existing drivetrain design workflow at Volvo Cars is explored