27716 research outputs found
Sort by
The Impact of Updated Reaction Rates on 26Al Production and Destruction in Very Massive Stars
Get PDFMachine learning implementation in electrical machine modelling and fault diagnosis
No full textMachine-learning (ML) particularly deep neural networks (NN) have been developed rapidly over the past decade, enabling powerful new tools for modelling, design, prediction, and fault diagnosis in electrical machines. This thesis explores how ML can enhance the modelling and fault diagnosis of permanent-magnet synchronous machines (PMSMs).
Firstly, this thesis presents a high-fidelity PMSM model that combines multiple lightweight NN blocks with established physical relationships. The framework captures magnetic saturation, spatial harmonics, temperature effects and iron losses without relying on large look-up tables (LUT) or simplifying assumptions that limit operating range. Compared with conventional finite-element or LUT-based approaches, the proposed model achieves comparable accuracy while reducing memory requirement and computational load.
The second part of the work focuses on open-circuit (OC) fault diagnosis in PMSM drives. Traditional model-based or signal-analysis methods are either parameter-sensitive or computationally intensive. This thesis introduces ‘normalised space-vector current sorting’ to convert three-phase currents into distinctive sequence-like patterns, which are then classified by a compact one-dimensional convolutional neural network (1D-CNN). Extensive offline tests, robustness analyses and real-time experiments confirm high diagnostic accuracy across a wide operating envelope.
Given that labelled fault data are scarce and simulation data suffer from domain shift, two simulation data-based strategies are investigated. (i) Data augmentation + multiscale 1D-CNN: prior knowledge based synthetic perturbations broaden the training set, and receptive-field scaling helps the network learn features that generalise to unseen real-world signals. (ii) Metric-learning with TripletNet, a neural-network framework that learns discriminative feature embeddings by comparing triplets of samples (anchor, positive, and negative): this structured embedding space improves separability even with few labels. Both approaches are validated in target experiment dataset.
Finally, this thesis also investigates the bearing impedance modelling to aid the prediction of high-frequency bearing currents in voltage source inverter-fed machines. A probabilistic ML framework is developed to model bearing impedance and its transition from predominantly capacitive to resistive behaviour. The proposed bearing impedance model is a conner stone for an accurate bearing current prediction, ultimately improving the bearing lifetime of electrical machines
Distributing Music in Space: Exploring the Impact of Spatial Directionality on Musical Material in a Composition Portfolio of Instrumental/Vocal Music
Get PDFThis research focuses on exploring how the spatial division of musical material can affect the perception and interpretation of music. Specifically, the study examines how allocating fragmented musical material to different musicians positioned around an auditorium can alter the listener’s overall experience. To investigate the questions of how the spatial distribution of musical material affects its perception in terms of distance and directionality and what compositional strategies can be employed to enhance the spatial experience for listeners in a live performance setting, I have created a portfolio of instrumental/vocal compositions experimenting with different approaches to distributing musical material in space.
Each of the pieces explores a different arrangement of musicians to create a spatially dispersed sound. Through this arrangement, I aim to examine the impact of the spatialisation of musical elements, focused more on their directional orientation (though I also consider distance), on the listener’s perception of various musical features, including structure, timbre, texture, rhythm, and harmony.
The research may contribute not only to music composition and performance but also to aspects of spatial sound design. By clarifying how spatial distribution shapes musical material, the study offers techniques for composing and performing instrumental/vocal music that is sensitive to its spatial context. Ultimately, it seeks to enhance listener engagement by opening new avenues for immersive and dynamic musical experiences. Practically, these findings can inform repertoire planning, staging and rehearsal strategies for ensembles that use spatial placement as a perceptual structuring device in live performance
The Royal Progresses of James VI & I, c.1580-1625: Negotiating Space and Performing Politics on the road.
No full textBeyond the Canvas: Translations Across Media and Dimensions, the Popular Canon and 'Printorial' Reform in the Long Nineteenth Century
No full textAcross the long nineteenth century, image reproductions acted as sites of experimentation that were key to establishing art history as a discipline. Chains of reproductions developed as images were repeatedly copied and translated into a wider range of media than ever before. Though widespread, this expansion of images into multiple media has received limited exploration, often in favour of reproductions into traditional printed, painted and drawn media. Focusing on reproductions where images were translated into relief formats, this thesis identifies a new ‘printorial’ genre that expanded the prominent graphic pictoriality of prints to a wide range of reproductive objects. Imagery was translated to be understood beyond the canvas through new media and contexts that exemplified the nineteenth-century preoccupation with reform, with a focus on developing aesthetic skill and knowledge.
This study works at the intersection of print, sculptural, reception, art theoretical and material culture studies to explore how ‘printorial’ objects reformed visual and material culture through widening popular access to imagery. In working across reproductions in different media, scales and dimensions, this thesis challenges the tendency to prioritise ‘high art’ mediums and forms connections between women, educational projects, amateur productions, industrial firms and professional artists. This methodology reveals a much fuller understanding of which images were most frequently reproduced and formed popular canons for audiences, often in contrast to that promoted by academic authorities. Through exploring a holistic view of the intersections of intermediality, secular and religious contexts, and high and low culture, this thesis highlights the hitherto unrecognised, but fruitful, relationship between material experimentation and the translated dissemination of a popular visual canon. Exploring the realms of creating and displaying reproductions, and how viewers might interact with these, evidences the versatility and expanse of ‘printorial’ reliefs as a mode of visual reform in the long nineteenth century
Advanced Machine Learning Approaches for Comprehensive Cardiovascular Disease Risk Prediction Using Synthetic Data and Dynamic Feature Selection
No full textCardiovascular diseases (CVD) are a leading cause of global mortality, highlighting the need for accurate and reliable risk prediction models. Traditional CVD risk assessment tools, such as Framingham, SCORE, and QRISK, have several limitations that affect their accuracy and applicability. These tools typically focus on a narrow set of major risk factors, potentially overlooking important non-traditional factors, resulting in a less comprehensive risk assessment. Additionally, they often rely on linear models, which may fail to capture complex, non-linear interactions within the data. This thesis addresses the limitations of traditional CVD risk assessment tools by developing a hybrid predictive framework that integrates advanced machine learning (ML) techniques to enhance the accuracy of Coronary Artery Calcium (CAC) score prediction and CVD risk assessment using both traditional and non-traditional risk factors. The research is structured around three key objectives: generating synthetic data, enhancing feature selection, and developing a hybrid approach. To address data limitations, a Tabular Generative Adversarial Network (GAN) was enhanced to generate high-quality synthetic data, effectively expanding the training dataset and improving model robustness. Feature selection was further refined through an adaptive SHAP-based
method, which dynamically adjusts feature importance thresholds to capture both traditional and non-traditional CVD risk factors more accurately. Finally, a hybrid approach combining hyperparameter tuning algorithms (Genetic Algorithms, Particle Swarm Optimisation, and Bayesian Optimisation) with Gradient Boosting algorithms (XGBoost, LightGBM, and CatBoost) was implemented to maximise predictive accuracy. This two-stage model first predicts CAC scores and then uses these predictions, alongside additional risk factors, to assess the likelihood of CVD events. Results demonstrate that the hybrid approach consistently enhances prediction accuracy across multiple metrics, with the CatBoost model particularly outperforming in both CAC score prediction and CVD classification
Gastrostomy procedures in England: current practices, patient outcomes and understanding of the associated decision making by healthcare professionals
No full textIntroduction
A gastrostomy can be life-extending; however it is also associated with an increase in morbidity and mortality. This research aims to describe current gastrostomy practices and associated mortality, and understand the factors involved in the gastrostomy decision-making process.
Methods
1. A retrospective study using the national endoscopy database (NED) and a national survey to understand clinical practice around gastrostomies.
2. Analysis of multicentre hospital records to determine how 30-day gastrostomy mortality varies and assess the validity of the Sheffield gastrostomy score (SGS).
3. Evaluation of factors influencing gastrostomy-tube removal.
4. Semi-structured interviews with doctors to understand what influences gastrostomy decision-making.
Results
1. The survey response was 69%. There was a 60% reduction in percutaneous endoscopic gastrostomy (PEG) insertions compared to previous estimates, this was validated with NED (≈6,500 vs 17,000 in 2010 (p<0.01)). Survey responses also revealed that consultant involvement in gastrostomy decision making, MDTs and aftercare provision had significantly increased.
2. PEGs =5% 30-day mortality, radiologically inserted gastrostomies (RIGs) =5.5%, per-oral image guided gastrostomies (PIGs) =7.2% (p=0.215). SGS was validated in PEGs, RIGs and PIGs. (n=1977)
3. Gastrostomy removal was associated by age, residence and head and neck cancer (p<0.05). 31.2% of gastrostomies were removed within 3 years. (n=251)
4. Eighteen doctors were interviewed and 4 themes derived: Strengths and weaknesses, process and influences, problems, and factors optimising decision-making.
Conclusions
This thesis shows that there are still areas that can be improved in gastrostomy selection and emphasises the importance of discussions amongst colleagues, patients, and families. Data suggests that MDT discussions have improved but do not occur for everyone. Information on mortality, including validated scoring systems, may be useful in the consent process. Some hospitals still lack effective referral systems for gastrostomy insertion, having these correctly in place should improve the selection process
Applying the COM-B model of behaviour change to cervical screening attendance in young women.
Get PDFCervical cancer, which is caused by the human papillomavirus (HPV), results in around 3,000 new cancer cases yearly in the UK. Cervical cancer rates in the UK have increased in young women over the last decade, and screening attendance has fallen to a 10-year low. As the majority of women and people with a cervix now reaching the screening age (24.5 years old) will be HPV vaccinated, research is needed to assess the impact of this successful immunisation programme on screening behaviours and further our understanding of the current barriers and facilitators to screening for both attendees and non-attendees.
In a systematic review of 106 studies looking at barriers, facilitators and factors associated with cervical screening in young women, it was found that there was an overall lack of application of theoretical models in cervical screening research. Therefore, I applied the COM-B model to a two-stage reflexive thematic analysis in a qualitative study of semi-structured interviews. This provided a more in-depth insight into the current barriers and facilitators to cervical screening. Further, a cross-sectional study found reflective motivations to be the only significant predictor of cervical screening. Using this evidence, an infographic was designed targeting motivational factors. The intention or motivation to attend cervical screening did not significantly improve. It is possible that the intervention infographic was not effective due to ceiling effects. There was some evidence to suggest that the infographic would be useful for those with neutral or negative baseline intentions to attend cervical screening.
Overall, the work in this PhD suggests that targeting reflective motivations is key to improving cervical screening attendance in young women and people with a cervix, particularly those with lower initial intentions, and that leveraging the positive impact of HPV immunisation through education can further increase uptake
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇