17628 research outputs found
Sort by
Energy market connectedness: A tale of two crises
Full text linkWe investigate connectedness within and across oil, coal and natural gas markets during the COVID-19 pandemic and the global energy crisis (GEC) using a time-varying parameter vector autoregression model and multiple price benchmarks. Our findings show that total connectedness spiked sharply but briefly during the pandemic and rose persistently during the GEC, reflecting crisis-specific developments that impacted demand and supply. At the aggregate level, spillover patterns varied across crises except for coal, with oil (natural gas) weakening (strengthening) as a transmitter during the pandemic, with the reverse observed during the GEC. The influence of specific benchmarks shifted in response to geopolitical events and policy interventions. We identify uncertainty, stock market dynamics, sentiment and energy market innovation as drivers of energy market connectedness. We further propose and test a trading strategy based on the strength and growth of pairwise connectedness, which yields profitable out-of-sample results. Our study contributes to the literature by offering a multi-benchmark analysis of fossil fuel market connectedness during two major crises, with the results providing insights and implications for market participants and policymakers
A Hermite Block-pulse hybrid method for nonlinear fractional optimal control problems
No full textFractional-order models capture memory and hereditary effects that cannot be represented by classical differential equations, yet their non-local nature makes associated optimal control problems computationally demanding. We propose a framework for solving fractional optimal control problems (FOCPs) by synergizing hybrid Hermite and Block-pulse functions with operational matrix techniques. The proposed methodology constructs a hybrid basis that combines the spectral accuracy of Hermite polynomials with the temporal localization of Block-pulse functions, enabling efficient discretization of fractional order dynamics. We derive rigorous operational matrices for Riemann-Liouville fractional integration and multiplication within this hybrid framework. A thorough convergence analysis shows exponential local and algebraic global error decay. These matrices transform FOCPs into purely algebraic optimization problems through generalized Lagrange multipliers, effectively handling both Caputo derivatives and non-smooth control inputs. Numerical case studies illustrate the method’s applicability to nonlinear FOCPs, demonstrating competitive computational accuracy compared to existing approaches
Care professionals’ perceptions of the use of PainChek ® among people living with dementia in Scotland
Full text linkBackground: Moderate to severe dementia is characterised by cognitive decline, communication challenges, and an increased risk of comorbidities, including chronic pain, which is a critical aspect of care. In Scotland, it is estimated that 90,000 people are diagnosed with dementia, 3,000 of whom are aged under 65 years. The Scottish government has been exploring innovative technologies to improve the assessment and management of pain in dementia care. Methods: We conducted a descriptive online survey with care staff to elicit their perceptions on the use of the digital tool PainChek®, for the assessment and management of pain in Scottish care homes as part of a wider evaluation. The study sought to ascertain the outcomes of the digital tool and novel pain assessment strategy, its impact on care, its potential for wider implementation, and the barriers and facilitators to its use. Twenty-eight care home staff who had used PainChek® completed an online survey administered via Microsoft Teams. Results: PainChek® was perceived by respondents as a user-friendly and effective tool for improving pain assessment and management, especially for non-verbal individuals. Staff reported improved decision-making, more accurate assessments, and a greater capacity to provide person-centred care. Reported facilitators for the adoption of PainChek® included strong management support, comprehensive training, and adequate resources. While respondents generally viewed PainChek® positively, staff were uncertain regarding its related costs. Conclusions: PainChek® has the potential to improve pain assessment and management in Scottish care homes. Staff highlighted enhanced decision making, more precise pain assessment and its user-friendliness. Its compatibility with national digital strategies and electronic health records could enhance personalised pain management for individuals with dementia across Scotland. We, therefore, recommend continued engagement with staff, providing clear information on how PainChek® can boost efficiency and improve patient-centred care to encourage the continued use of the digital application
Quantification and Characterisation of Microplastics in Organic Waste-Derived Soil Amendments
No full textThe spread of organic waste-derived amendments such as compost, anaerobic digestate and biosolids has been identified as a major source of microplastics in agricultural soil, with potential negative environmental and human health effects. Due to lacking regulatory frameworks and standard monitoring procedures, the extent of microplastic contamination in Scottish organic waste-derived soil amendments is still poorly understood. This study investigated the presence, quantity and characteristics (morphology, density and colour) of microplastics in anaerobically digested biosolids, green-waste-derived compost and food-waste-derived anaerobic digestate produced in Scotland. Microplastics (100—5000 µm) were present in all analysed samples in concentrations ranging from 34 to 160 particles g−1 dw, with the highest levels found in biosolids, followed by digestate and compost. High-density fibres represented 55.8—66.4% of microplastics in biosolids, likely polyester from the domestic washing of textiles. In addition to microplastics, > 20,000 cellulosic microfibres g−1 dw, likely textile-derived natural fibres, were detected in biosolids, and were absent in other samples. Microplastic fibres of a wider-density-range represented 72% of microplastic in compost, while high-density microplastic fragments (34%) and fibres (24%) were the most abundant microplastics in digestate. Based on the results, it was estimated that compost, anaerobic digestates and biosolids could respectively introduce 3.17 × 1012, 5.9 × 1011 and 7.2 × 1012 microplastic particles measuring 100—5000 μm to Scottish land, annually. These findings highlight the extent of microplastic contamination in terrestrial environments across Scotland, underscoring the need for standardised routine monitoring, enhanced waste management practices, and stricter regulatory measures
Integrated Biomimetic 2D-LC and Permeapad® Assay for Profiling the Transdermal Diffusion of Pharmaceutical Compounds
No full textA comprehensive two-dimensional liquid chromatography platform (LC × LC) was developed and validated for dermal permeability studies. In this implementation, the two separation dimensions were applied to mimic the layered structure of human skin: a ceramide-like stationary phase in the first dimension (1D) to simulate the lipid-rich epidermis, and an immobilized artificial membrane (IAM) phase in the second (2D) to emulate the dermis. Experimental conditions were optimised to reflect the microenvironment of the in vivo skin. For validation purposes, 43 pharmaceutical and cosmetic compounds whose transdermal permeability coefficients (log Kp) were known from the scientific literature were selected as model solutes. A good degree of separation was achieved across the whole dataset, and affinity profiles correlated with transdermal passage properties, suggesting that retention within specific chromatographic ranges may be predictive of skin permeation. To complement this approach, mass diffusion measurements were also conducted using Permeapad® 96-well plates and LC was performed on a narrow bore column in MS-friendly conditions. These log Kp values were compared against both in vivo and chromatographic retention data. The combined use of these techniques offers a strategic framework for profiling new chemical entities for their dermal absorption in a manner that is both ethically compliant and eco-sustainable
Audio-visual speech-in-noise tests for evaluating speech reception thresholds: A scoping review
No full textObjective: To evaluate the advancements in speech intelligibility testing over the recent decades, with a particular emphasis on the development of audiovisual speech in noise tests that incorporate both auditory and visual modalities for the measurement of speech recognition thresholds. Design: A scoping review was conducted systematically to examine the existing literature on speech intelligibility testing methods. Following comprehensive screening process, studies were selected for detailed analysis, focusing on audiovisual integration and potential for remote or automated administration within studies methodologies. Study Sample: The review encompassed 11 scholarly articles that investigated diverse approaches to speech intelligibility testing. Results: The analysis revealed variability in the accuracy and reliability of speech intelligibility testing methods. Although certain methods demonstrated efficacy in incorporating audiovisual cues, none of the reviewed studies included provisions for remote administration, thereby necessitating the presence of a clinician for test execution. This limitation underscores the imperative for further research development of remote testing methodologies that leverage audiovisual technologies to assess speech in noise. Conclusions: The findings of this review underscore the critical need for advancement in speech intelligibility testing methodologies particularly integrating audiovisual components and enabling remote administration. The development in this domain holds significant potential to enhance the assessment and implementation of assistive technologies for individuals with hearing impairments
Malware recognition using novel convolutional neural network with residual connections
No full textWith the evolving use of IoT devices, the accurate recognition of different types of malware is critical for effective threat mitigation in cybersecurity. There is a significant amount of research on malware recognition; however, the existing state-of-the-art approaches are far from perfect. As IoT networks expand and become more widespread, the challenges associated with malware recognition are becoming more frequent and difficult to address. To this end, this paper presents a novel convolutional neural network architecture (CNN) with residual connections to efficiently recognize malware by analyzing the image representations of malware files. The residual connections in the proposed CNN architecture enable the network to capture complex patterns in the malware code and behavior at multiple scales. The proposed model does not require any input preprocessing or data augmentation, underscoring the model’s adaptability and ease of integration into existing image classification tasks. The model is evaluated on three widely used datasets, namely Malimg, Malevis, and APTMalware, highlighting its effectiveness and robustness in distinguishing a variety of malware classes. To demonstrate the model’s robustness and generalization ability, we further evaluated it on the APTMalware dataset, which represents real-world advanced persistent threats. This addition validates the model’s applicability in modern cybersecurity environments. Additionally, the proposed model is benchmarked against a diverse set of deep learning architectures, including EfficientNetB0, ResNet50, Inception V3, EfficientNetV2, and DenseNet121, providing a well-rounded comparative analysis with recent high-performing models. The experimental results establish the superior performance of the proposed CNN network with residual connections, achieving a 10-fold cross-validation accuracy of 99.46% on the Malimg dataset and 95.10% on the Malevis dataset, while also showing competitive results on the APTMalware dataset, highlighting the robustness of the proposed model across different malware types
How to … Maximise Your Clinical Education Research Impact
Full text linkResearchers are increasingly expected to demonstrate that their work leads to impact beyond academia, particularly when seeking research funding. Yet practical guidance on how to plan for, achieve and evidence impact in Clinical Education Research (ClinEdR) remains limited. This guide offers pragmatic advice to support postdoctoral researchers in designing fundable projects and producing work that leads to meaningful change. We argue that impact should be considered from the start of a research project, not added at the end. Early and ongoing engagement with potential research beneficiaries, including educators, learners, institutions, professional bodies, policymakers and patients, helps ensure that research addresses real-world priorities. Such engagement can shape research questions, study design and outputs, strengthening both grant applications and the likelihood of downstream uptake. This how to guide outlines the value of articulating a clear impact goal and mapping a credible pathway to impact that distinguishes between outputs, outcomes and longer term benefits. We discuss how methodological choices can enable impact, with attention to participatory, design-based and implementation-focused approaches that have change embedded within them. We also highlight the importance of embedding impact checkpoints throughout a project, recognising that impact is often non-linear, delayed and context dependent. Finally, we describe deliberate, audience-specific dissemination strategies and practical ways to evidence impact beyond traditional academic metrics. Embedding impact thinking across the research lifecycle can enhance the relevance, reach and value of ClinEdR, while supporting postdoctoral researchers to meet funder, institutional and societal expectations
Well-posedness of Lur'e systems with feedthrough
No full textFor a large class of Lur'e systems with time-varying nonlinearities and feedthrough we consider several well-posedness issues, namely:existence, continuation, blow-up in finite-time, forward completeness and uniqueness of solutions. Lur'e systems with feedthrough are systems of forced, nonlinear ordinary differential equations coupled with a nonlinear algebraic equation determining the output of the system. The presence of feedthrough means that the algebraic equation is implicit in the output, and, in general, the output may not be expressible by an analytic formula in terms of the state and the input. Simple examples illustrate that the well-posedness properties of such systems are not necessarily guaranteed by assumptions sufficient for the corresponding well-posedness properties of Lur'e systems without feedthrough. We provide sufficient conditions for the well-posedness properties mentioned above, using global inversion theorems from real analysis and tools from non-smooth analysis and differential inclusions. The theory is illustrated with examples
UHF Partial Discharge Detection for Power Transformer Based on Hilbert Fractal Antenna with Different Feeding Techniques
Full text linkIn power transformer, Partial Discharge (PD) can cause a major equipment risk and lead to further accidents. One of the PD detection techniques is Ultra-High Frequency (UHF) antenna. UHF antenna have a major advantage which is it immune to external interference. However, the choice of feeding technique used to connect the antenna to the transmission line can limit the performance of UHF antenna. This paper was conducted by comparing microstrip line feed and coaxial feed for Ultra-High Frequency (UHF) PD detection. The antenna is aimed to operate in the PD frequency range (300-3000 MHz) and there is no specific frequency band that PD radiate in a measurement. The frequency of UHF PD measurements in power transformers varies based on the individual case and is impacted by several factors, including the PD source and locations of PD in the transformer. In this work, 4 th order Hilbert fractal antenna are selected as patch antenna for UHF PD detection. The proposed design for the antenna based on the appropriate size for internal installation in the power transformer. All these results showed that either microstrip line feed or coaxial feed technique are able to capture the PD in the range of UHF for the power transformer. However, the coaxial feed antenna is preferable as it shows that the fabricated measurement is consistent with the simulation and had a higher number of resonant frequencies
Data sources:
Access Repository Dashboard
Do you manage Repository@Napier? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!