Publikationer från Uppsala Universitet
Not a member yet
127603 research outputs found
Sort by
To site or not to site : A framework development for battery energy storage system siting using data-driven insights
No full textThe objective of this study was to develop a structured framework for identifying suitable locations for battery energy storage systems (BESS) in Germany. The framework was constructed using insights from literature, German grid information and expert dialogs. The framework combines technical, economic and spatial factors relevant for BESS siting. To test the practical use of the framework, a case study was conducted based on the 110 kV grid in northeast Germany. A simplified power grid model was constructed in Python using PandaPower and included wind power injection, regional demand and three potential BESS locations. The framework focused on evaluating how relevant each identified siting factor is and how the factor could be applied using publicly available open data. By mapping the importance of each factor against its data availability, a four-field analysis was performed to determine which factors are most suitable for inclusion in an initial screening tool. The result of the framework shows that residual energy, proximity to substations and land suitability was the factors with the highest relevance and most available data. The case study shows that Lubmin and Greifswald perform similarly well as BESS sites, while Wolgast shows significantly lower total usage and is less suitable for BESS placement. The study concludes that the framework enables initial BESS siting selection based on the identified factors and open data. Further development could integrate economic factors, larger regional data set of Germany and automated decision making.
Mitigating Covariate Shift in Machine Learning Based Prediction of Intracranial Hypertension
No full textIntracranial hypertension is a critical complication following a traumatic brain injury, necessitating early intervention to improve patient outcomes. While machine learning has shown potential in predicting intracranial hypertension, real-world deployment may be challenging due to covariate shift. Moreover, patients pose additional risk due to alarm fatigue. As limited research has been conducted on the subject, this study aimed to analyze the effects of dataset subsetting and covariate weighting on mitigating covariate shift in intracranial hypertension prediction. To investigate this, a debouncing mechanism was developed based on insights from domain experts, which revealed that over 92% of the triggered alarms were deemed clinically irrelevant, suggesting the results from previous studies may be skewed and not directly applicable to real-world deployment scenarios. Three different models were therefore developed to improve alarm performance: (1) a baseline model trained on all patients, (2) a subset model trained only on patients without ongoing intracranial hypertension, and (3) a weighted model trained on all patients, incorporating covariate weighting by kernel mean matching with a time warp edit distance-based kernel. The results showed that neither strategy improved model performance, in fact, subsetting worsened it. While the weighted model produced comparable results to the baseline model, the computational cost of adjusting temporal misalignments between intracranial pressure signals was substantial. These results emphasize the need for evaluation frameworks that are adapted to assessing real-world alarm performance, as well as scalable strategies for mitigating covariate shift
Teachers’ administrative burden and job satisfaction
No full textThis thesis investigates if there is a connection between teachers' administrative workload and job satisfaction using an econometric analysis at the micro (teacher) and macro (country) level. This is done by utilizing the Teaching and Learning International Survey (TALIS) conducted in 2013 and 2018. The study focuses on lower secondary school (ISCED level 2) and includes 177 525 observations scattered across 29 countries and two economic subregions. Findings reveal that job satisfaction within the categories Career Choice, School and Work Environment, and Overall Job Satisfaction declined as the proportion of administrative hours increased. While the study successfully established a negative relation between administrative tasks and different dimensions of job satisfaction, it is at the moment not ascertained to be causal in nature
Breaking with the Krona: A Counterfactual Analysis of Sweden’s Post-COVID Inflation Shock
No full textThis thesis examines how Sweden’s decision to retain its own currency (the krona) influenced its economic outcomes during the post-COVID inflation shock, by constructing a counterfactual scenario in which Sweden had adopted the euro. We employ the Synthetic Control Method to create a synthetic Sweden, a weighted composite of eurozone economies. This allows us to estimate how key macroeconomic indicators, specifically real GDP per capita, would have evolved under euro membership. The analysis finds that in the aftermath of the 2021 onwards inflation shock, the euro-based synthetic Sweden exhibits a faster recovery, narrowing the gap in real GDP per capita. Key findings indicate that while Sweden’s GDP per capita remained higher with the krona than it would have with the euro, the euro scenario mitigated some effects of the inflation shock, leading to a more rapid post-shock rebound. These results highlight a contingent trade-off: the krona conferred advantages in long-run growth and internal crisis response, whereas euro membership could have provided greater stability during an extreme inflationary episode
Kan vi skjuta upp det? : Frågan om miljömässigt hållbar digitalisering av kulturarv
No full textThe escalating effects of climate change are becoming a discernible threat to tangible cultural heritage, and to preserve the heritage at risk digital preservation has become a common practice by institutions. However, paradoxically, the ICT (Information & Communication Technology) infrastructure and cloud services used in digitalisation is simultaneously contributing to the escalation of climate change. This thesis advocates for the significance of Environmentally Sustainable Cultural Heritage Digitalisation, abbreviated ESCHD, by investigating the triadic relationship between the fields of environmental sustainability, digitalisation and cultural heritage. This is conducted through a quantitative gap analysis of identifying key words in cultural heritage policy documents, as well as a qualitative interview study with professionals working in one or several of the ESCHD fields. The research is done from a Swedish perspective with a European influence, and the results are supported using a theoretical framework containing key concepts surrounding inaction. The gap analysis was carried out on 16 documents which identified that only a third contained instances of environmental sustainability. The remaining documents contained vague mentions, or in some cases, no mentions at all. The thematical analysis of the six interviews revealed three overarching concerns: (1) the presence of a knowledge gap, (2) the point of responsibility, and (3) goal conflict. Although there is ambition within the sector to work more sustainably, action is hampered by issues such as a lack of awareness or knowledge, access to supporting data, the reluctance to overstep the boundaries of one’s own field, and the organisational prioritisation of other goals e.g. economical or security concerns. The findings highlight the importance of further research within the ESCHD field, emphasising the urgency of collective action. This is a two-year master’s thesis in Museum and Cultural Heritage Studies. De eskalerande effekterna av klimatförändringarna har blivit ett märkbart hot mot materiellt kulturarv, och för att bevara det hotade kulturarvet har digitalt bevarande blivit en vanlig praxis bland institutioner. Paradoxalt nog bidrar dock informations- och kommunikationsteknikens infrastruktur och molntjänster som används i digitaliseringspraxis till eskaleringen av klimatförändringarna. Denna masteruppsats uppmärksammar vikten av miljömässigt hållbar kulturarvsdigitalisering (förkortat till ESCHD) genom att undersöka det trefaldiga förhållandet mellan områdena miljämässig hållbarhet, digitalisering, och kulturarv. Detta görs genom en kvantitativ gap-analys för att identifiera nyckelord i policydokument från kulturarvssektorn, samt en kvalitativ intervjustudie med yrkesverksamma som arbetar inom ett eller flera av områdena inom ESCHD-fältet. Undersökningarna görs utifrån ett svenskt perspektiv med europeiskt inflytande, och resultaten stödjs av ett teoretiskt ramverk som centreras kring passivitet. Gap-analysen genomfördes på 16 dokument, och det identifierades att endast en tredjedel nämnde miljömässig hållbarhet. De återstående dokumenten innehöll vaga eller i vissa fall inga omnämnanden alls. Den tematiska analysen av de sex intervjuerna visade på tre övergripande frågor: (1) förekomsten av ett kunskapsglapp, (2)ansvar, och (3) målkonflikt. Även om det finns en ambition inom sektorn att arbeta mer hållbart så hämmas agerande av problem som bristande medvetenhet eller kunskap, tillgång till stödjande data, ovilja att överskrida det egna områdets gränser, och organisatoriska prioriteringar av andra mål inom till exempel ekonomi eller säkerhet. Resultaten belyser vikten av ytterligare forskning inom ESCHD-fältet, och betonar den brådskande situationen och betydelsen av kollektiva åtgärder. Detta är en tvåårig masteruppsats i musei- och kulturarvsstudier.
Comorbid posttraumatic stress disorder complicates eating disorder treatment : A nationwide study
No full textIn eating disorders, it is unclear if the experience of a trauma alone is sufficient for the more frequent occurrence of binge eating, self-induced vomiting, self-harm, and suicidality or if these ensue primarily in those individuals who develop posttraumatic stress disorder (PTSD). Distribution appropriate regression analyses in the world's largest clinical eating disorder sample in Sweden (n = 8906) tested for associations between trauma or PTSD and 1) eating disorder type, 2) impulsive eating disorder behaviours, 3) non-suicidal self-injury, and 4) different forms of suicidality. Most variables apart from impulsive disordered-eating behaviours were clinician recorded. In Sweden, 16 % of patients had experienced trauma and 4 % had PTSD. Compared with anorexia nervosa restricting subtype, individuals with the binge-eating/purging subtype or purging disorder were significantly more likely to have comorbid PTSD. Comorbid PTSD at registration to clinical services was significantly cross-sectionally associated with more frequent self-injurious and suicidal symptoms and longitudinally with binge eating frequency, and self-injurious and suicidal symptoms at 12-month follow-up. Our results show no clear support for the experience of trauma alone having similar effects. This underscores the importance of assessing traumatic experiences leading to PTSD in clinical settings as they represent treatment complicating factors
Approaching the Sensitivity Limit for Phenotypic Antibiotic Susceptibility Testing in Escherichia coli
No full textSepsis, when resulting from bacterial bloodstream infections (BSIs), demands rapid and precise antimicrobial treatment to reduce mortality. However, current diagnostic workflows for antibiotic susceptibility testing (AST) rely on time consuming blood culturing, delaying targeted therapy. This thesis contributes to the development of phenotypic AST (pAST) at the single-cell level, aiming to bypass the culturing step and enable faster, more accurate treatment decisions directly from patient samples. This study investigates sequential antibiotic testing of individual Escherichia coli lineages within a sample, focusing on whether prior exposure to an antibiotic to which the bacterium is originally resistant affects its susceptibility to a subsequent antibiotic in the treatment sequence. By combining microfluidics with time-lapse microscopy and computational analysis, this study monitored phenotypic responses, such as growth arrest and cell elongation, at the single-cell level over time. Ciprofloxacin (CIP) served as the first-line antibiotic, followed by second-line antibiotics including amikacin (AMI), tigecycline (TIG) and cefotaxime (CTA). The results indicate that a strains prior exposure to CIP, despite resistance to the antibiotic, did not affect susceptibility to the subsequent antibiotics tested. These findings show that single-cell sequential pAST is possible and the work contributes to the development of rapid diagnostic strategies for BSIs that are independent of blood culturing
Development and Validation of a Geant4 Monte Carlo Model for Excillum’s NanoTube Transmission X-Ray Source
No full textAccurate modeling of spot size and beam characteristics in X-ray sources is crucial for opti-mizing design and ensuring high performance. The study investigated the use of the Monte-Carlo simulation toolkit Geant4 for simulating the NanoTube, one of Excillum’s transmis-sion X-ray tubes, and validating the simulations by comparing simulated and experimentallymeasured X-ray spots. Simulated and measured spot shapes agreed well with only small dis-crepancies and an optimization of the target thickness was done to find a thickness whichminimizes the tails and full width half maximum of the spot. The computational efficiencyof the software was compared with another Monte-Carlo code, PENELOPE, already in useby the company, and it was found that Geant4 had around one magnitude higher efficiencyusing the current setup. Further improvements can be done to achieve faster simulationsby incorporating more efficient variance reduction techniques, optimizing step sizes as wellas the possibility of accelerating simulations with GPUs. These findings demonstrate thatGeant4 can both reproduce the NanoTube’s spot geometry and shorten simulation time,enabling more rapid design iterations for next-generation transmission tubes
Process control development of apatite leaching
No full textThis thesis describes the design and implementation of a process control system for apatite leaching, a critical step in LKAB’s efforts to recover phosphorus and rare earth elements from mining by-products. The work was conducted in collaboration with ABB and aimed to replicate, on a laboratory scale, the processes intended for a future full-scale hydrometallurgical plant in Luleå, Sweden. Using ABB’s 800xA distributed control system, a modular automation architecture was developed to enable flexible and scalable operation. The project involved selecting suitable field equipment, designing control logic in Control Builder, and creating operator interfaces with Graphics Builder. Special focus was given to dynamic I/O addressing, simulation via virtual controllers, and integration of PID-based flow regulation. The system was validated through a combination of partial system simulations and physical tests, including temperature measurement and pump control. The resulting control system successfully measured and regulated temperature and pump speed for the apatite leaching stage, and provides a robust foundation for integrating subsequent process steps such as arsenic precipitation and acid extraction. Although integration of feeder control and weight reading remains in development, the system forms a solid base for future expansion.
