30793 research outputs found
Sort by
State-enhanced attention network for optimisation of energy and yield in gas atomised metal powder production
Get PDFGas atomisation is a widely used technique for producing spherical metal powder feedstock for additive manufacturing. However, the process parameters suffer from variability and inefficiency in balancing powder yield, energy consumption, and particle size distribution. Optimising these complex, interdependent parameters pose a significant challenge. This work proposes a novel State-Enhanced Attention Network architecture in a framework that simultaneously optimises yield and energy consumption during nitrogen gas atomisation for sustainable metal powder production. The novelty lies in integrating processed long-term memory states with the attention mechanism, enabling nuanced attention weighting. This allows the model to leverage global sequence context and recent state information for improved yield and energy predictions. The proposed network is trained and integrated into a non-dominated sorting genetic algorithm to enable multi-objective optimisation. This framework evolves a set of Pareto-optimal solutions that balance trade-offs between maximising yield and minimising energy consumption. The approach is evaluated using augmented real-world data from an industrial gas atomisation plant. The proposed model demonstrates significantly improved predictive accuracy on real-world datasets, compared with baseline deep learning models. Results highlight the capabilities of the proposed technique for automated, data-driven optimisation of gas atomisation, simultaneously improving yield, energy efficiency, quality control, and sustainability. The integrated deep learning and evolutionary optimisation framework also provides an innovative solution for enhanced control of additive manufacturing powder production processes.UK's Department of Energy, Security and Net Zero, under the Industrial Energy Transformation Fund (IETF22034)
Healing the split: An autoethnographic exploration of psychosis following trauma
Get PDFThis thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University LondonThis research project aims to explain how a trauma can lead to a false and distorted understanding of reality and how this can in turn lead to psychosis. Having worked for many years trying to understand her psychosis, the author wanted to research and explain what had happened to her. To fulfil this aim, an auto-ethnographic qualitative method was chosen, whereby the author utilised and analysed her own diary entries over time. This method allowed the researcher to disclose and explore aspects of her life that could have been too intrusive, or even potentially distressing had they been probed as part of a third-person perspective, the researcher becoming a participant in someone else’s project.
Psychosis is a condition characterized by subjective difficulties with reality. Winnicott’s explanations of the formation of the false self in psychosis are used to explore how reality’s understanding fails. Special consideration is given to Winnicott’s final paper “Fear of breakdown”, in which he indicated the existence of a “not lived” trauma in psychosis, because of ego immaturity and the subject’s inability to encompass the experience. The trauma needs to be remembered and lived through to resolve the psychosis. Bion’s theory of thought complements Winnicott’s thinking, with its explanation of the difficulties to process and the need to digest trauma. Ferenczi’s work is also utilized to evince how people may be led to behave in uncharacteristic manners, at times even displaying violent behaviour because of trauma. In this respect, the idea of ‘possession’ by another is put forward to interpret psychosis and its resulting behaviour. In addition, Bollas’s understanding of psychotic symptoms is looked at and especially how they can be brought to some form of resolution by helping the individual to integrate what is being externally projected.
Owing to the trauma experienced by the researcher, she failed to understand what was happening when the behaviour of her attacker forced her to internalize it with a different meaning of its reality. Her auto-ethnographic case is examined and compared, using the qualitative method of Thematic Analysis, to the published clinical cases and memoirs of Renée (Marguerite Sechehaye) and Marie Cardinal. This comparison shows how different traumas at different ages will have a different impact and different consequences on the individual, insofar as a trauma experienced by an adult will have less impact on cognition or thought forming, for instance, if before the trauma there was a healthy psychological development. However, psychosis following trauma remains a common factor, because the true self is forced into hiding and a false self with a false understanding becomes dominant. It is argued that it may not be due to ego immaturity that the traumatic experience is not integrated within the psyche, but – as in the researcher’s own experience – that it may be the behaviour of the abuser (or similar factors) compelling and imposing the failure to ‘live’ the experience.
Finally, the resolution of psychosis is explained by allowing the subjective truth of the victim of trauma to be remembered, processed, understood, and eventually newly integrated as reality. This would allow for the failure in the understanding of reality to be overcome. To look for those areas where reality has not been understood is recommended as the foundation for treatment, while the therapist needs to remain aware of how a dormant psychosis can still be triggered there
Multiagent deep reinforcement learning-based cooperative optimal operation with strong scalability for residential microgrid clusters
Get PDFData availability:
The authors do not have permission to share data.With the rapid development of smart home technology, residential microgrid (RM) clusters have become an important way to utilize the demand-side resources of large-scale housing. However, there are some key problems in existing RM cluster optimization methods, such as difficult in adapting to the local observable environment and with poor privacy and scalability. Therefore, this paper proposes a multi-agent deep reinforcement learning (MADRL)-based RM cluster optimization operation method. First, with the aim of minimizing the energy cost of each residence while satisfying the comfort level of residents and avoiding transformer overload, the optimization scheduling problem of an RM cluster is described as a Markov game with an unknown state transition probability function. Then, a novel MADRL method is proposed to determine the optimal operation strategy of multiple RMs in this game paradigm. Each agent in the proposed method contains a collective strategy model and an independent learner. The collective strategy model can simulate the energy consumption of other RMs in the system and reflect its operating behavior. In addition, an independent learner based on a soft actor-critic (SAC) framework is used to learn the optimal scheduling strategy interactively with the environment. The proposed method has a completely decentralized and scalable structure, which can deal with continuous high-dimensional state and action spaces only requires local observations and approximations during training. Finally, a numerical example is given to verify that the proposed method can not only learn a stable cooperative energy management strategy but can also be extended to large-scale RM cluster problems. This gives the strong scalability and a high potential for practical application.This work was supported in part by the National Natural Science Foundation of China under Grant 52107108
Seen but not heard: the voice of women at work and the mediating role of culture
Get PDFData availability statement:
The authors confirm that the data supporting the findings of this study are available within the article.While there is now an extensive body of literature on employee voice behaviour in the global North, research evidence from the global South is limited. This has constrained our understanding of the barriers that female workers face in expressing their views and concerns in developing countries such as Nigeria. This article examines the cultural factors that shape female employee voice behaviour in Nigerian workplaces. Using a meta-synthesis of 52 semi-structured interviews and approximately 200 h of non-participant observation, we identify a high-power distance orientation and patriarchal norms as two cultural factors that contribute to gender imbalance in the workplace, making it difficult for female employees to express their opinions, suggestions, ideas, or complaints about important workplace issues. Our findings highlight a system of patriarchal hegemony and gender inequality that makes voice behaviour difficult for female workers. The findings also show that contextualised religious norms and teachings encourage silence among female employees. We provide valuable insights into the cultural norms that inhibit female employee voice behaviour in the Global South context
CCL-MPC: Semi-supervised medical image segmentation via collaborative intra-inter contrastive learning and multi-perspective consistency
Get PDFData availability:
The code will be made available upon publications.Semi-supervised image segmentation extracts specific regions and tissues by utilizing extensive unlabeled images and limited labeled images, which can alleviate the dependence on plenty of accurately labeled data. However, it is difficult to learn robust feature representations due to the potential noise in pseudo-labels caused by inefficient consistency learning, and poor class diversity in feature spaces. To address this issue, we propose a semi-supervised medical image segmentation method using Collaborative intra-inter Contrastive Learning and Multi-Perspective Consistency (CCL-MPC). First, we propose a collaborative intra-inter contrastive learning strategy that includes symmetric bidirectional contrastive learning and certainty-guided contrastive learning, to exploit the intrinsic differences between inter-image and intra-image feature representations. Second, we design a multi-perspective consistency learning strategy to improve the class diversity by utilizing a dual-branch network and two augmented views. Additionally, we dynamically partition the pseudo-label certainty area for auxiliary consistency learning to reduce the potential noise during the training process. Experimental results on the publicly available datasets demonstrate that CCL-MPC can achieve better segmentation performance than the state-of-the-art methods for semi-supervised medical image segmentation tasks. The source code is available at https://github.com/EmarkZOU/CCL-MPC.This work is partly supported by National Natural Science Foundation of China (Nos. 61861024, 62271296, and 62201334), Scientific Research Program Funded by Shaanxi Provincial Education Department (Nos. 23JP022, and 23JP014), and General Project of Key Research and Development Programs in Shaanxi Province, China, Social Development Area (No. 2022SF-105)
Making a CASE for the third space
Get PDFThis thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University LondonThe Thinking Science programme or CASE (Cognitive Acceleration through Science Education) is a
sequence of thirty lessons that seek to develop pupil reasoning in the context of secondary science. CASE is
based upon the work of Jean Piaget and Lev Vygotsky and was very popular being used by hundreds of
schools, both in the UK and around the world, during the 1990’s and the 2000’s. The lessons and ideas are
still in use today and serve as a lasting legacy for the authors: Philip Adey, Michael Shayer and Carolyn
Yates. This thesis will present an insider’s view of CASE from a teacher who has been using the lessons for
34 years. This perspective will allow the author to articulate an experienced teachers thinking regarding the
principles of CASE. Alongside this an analysis of Vygotskyan psychology, illuminated by Bhabha’s ideas,
will be used to present the concept of the third space as a model to visualise the small group talk taking place
in the CASE classroom.
Since the advent of CASE much of the language it used to describe cognitive development, such as
metacognition, has become common parlance in educational settings. This thesis brings a practitioner
perspective to bear on the principles of CASE and the pupil talk it seeks to generate. Seven CASE lessons
were taught to a secondary, Year 9, class over a period of six months and video recorded. The resulting
lesson plans and transcripts, of small group conversations, enabled the Vygotskian foundations of the CASE
programme to be explored. Bhabha’s notion of the third space was used to categorise talk in a way that is
reflective of the real lives of children and the ideas they carry with them into the classroom. This analysis
enabled an innovative, colour, coding system to be developed to illustrate and track the contours of
children’s talk. This thesis brings a sense of clarity regarding the use of third space theory in educational
settings and bridges a gap in this canon as it takes place in the mainstream secondary classroom within the
context of a core subject. As a result third space talk has now been described and categorised in a way that
allows it to be applied across subjects and in other settings such as special schools.
The Vygotskian concept of a child’s lived experience, perezhivanie, is used to illustrate a possible route for
teacher’s to guide the conversations towards the types of talk supportive of cognitive development. This
research uses the idea of a trigger to show how appropriate challenge can be used to engage learners and this
is uniquely illustrated through the use of a weather map representation of pupil talk. This serves to refresh
CASE and will support and new generation of teachers to better manage the dynamics of small group talk at
a time when the pressure for children, yet again, to be seen and not heard is growing.
The key findings from this study for classroom practitioners can be summarised using the following
headings:
a) Inclusion,
b) Professional choice,
c) Construction talk and,
d) Hope in the zone.
Inclusion points to the fact that pupil funds of knowledge need to be utilised, in the third space, so
that their life-times worth of experiences brought to the classroom are included and acknowledged.
This is true for all learners and the potential of this is supported by the fact that majority of the
pupils highlighted in the transcripts were categorised by the school as SEND.
Professional choice references the fact that choices are made in the planning and conduct of
lessons. To promote thinking teachers can integrate triggers in a way that engages pupils in
intellectual inquiry. Triggers can be included in the lesson conduct or in response to the
conversational turns taken in real-time. Similarly small groups can be orchestrated to develop
construction talk that develops pupils thinking within the two planes described by Vygotsky – the
social and the psychological. If managed utilising the third space then the external interactions with
others support and enhance the inner cognitive life of the individual.
Construction talk is a living vocabulary that verbalises the inner thought life and so is one that
must be inclusive of all forms of talk. The five talk codes reflect the real, honest, messy and playful
conversations taking place in classrooms today as adolescents engage with the cognitive challenges
around which CASE is built. The sensitive use of triggers can engage and guide pupils towards
forms of talk that activate their funds of knowledge and develop both ownership of their talk and a
metacognitive awareness of their thinking.
Hope in the zone – the third space is a theoretical construct that has been successfully used and
applied in a range of educational settings. It has been used here to illuminate construction zone
activity as a powerful time of developmental potential for children facing abstract concepts in
science. Hope in the zone is possible with thoughtful planning and the use of specific classroom
management techniques following the clear pedagogical guidelines for such techniques that are
provided within this thesis.
The outcomes of this research form a practical and easy to apply guide to the use of the third space
as a bridge between pupil knowledge and curricular content and serves to both update and
reposition CASE for a new generation of teachers in a way that is inclusive of all learners. The time
has come for the third space to become more mainstream curricular vehicle. Historically it has been
hard to find and was an idea heavy on metaphor yet light on practical examples of realistic teaching
advice. This thesis shows what it looks and sounds like in the classroom in a way that is inclusive of
all learners. C.S. Lewis said that children are not a distraction from more important work - they are
the most important work. The outcomes of this thesis will help teachers to carry out such work,
work that will outlive CASE and last long beyond my life in the classroom
A Social Network Analysis of Opportunistic Behaviors in Government R&D programs
Get PDFDrawing on transaction costs analysis, this study investigates the effect of two partner selection strategies in government R&D programs: selection based on dyadic relation and network reputation of candidate partners. While governments play a vital role in mitigating opportunistic behavior, direct intervention of governments can increase administrative burdens and decrease efficiency, leading to higher costs for the government. Building upon existing literature on relational and network theories, the research aims to provide insights on the role of partner-selection strategies as effective self-enforcing mechanisms on opportunism control. A simulation model is proposed to track long-term changes in network configuration and transaction costs under project uncertainties. The base model demonstrated that selection based on relations forms a more cost-effective partner network. The next step is to analyze how the transaction costs of these two strategies change on the project uncertainty
Echo state networks in forecasting chaotic dynamics and emergent universalities
Get PDFThis thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University LondonDeep learning has become a popular way to accurately recognise, classify, generate and
forecast data in various complex scenarios. It is commonly based on artificial neural
networks, a type of machine learning architecture inspired by the human brain, that
consist of multiple layers of neurons with non-linear activations. Their inside workings
can be considered as a ‘black box’ because it is difficult, if not impossible, to truly
understand “how” or “why” such networks work. Furthermore, the training relies on
gradient decent methods which are inherently difficult and computationally expensive.
Reservoir Computing has recently emerged as a new paradigm aimed to alleviate
such difficulties. In this thesis, we study its particular realisation, known as Echo State
Networks (ESNs), which show promise in many tasks, particularly in forecasting the
dynamics of chaotic systems.
The thesis will provide a detailed discussion of the ESN architecture, hyperparameters,
and implementation. We introduce a new performance metric that highlights the
networks maintaining small errors for the longest duration and use it in GridSearch to
optimise ESN hyperparameters. We study the statistical properties of the correlation
matrix that is formed during training, offering a novel approach not applied to ESNs
or other types of recurrent networks. Our extensive numerical studies reveal universal
results, consistent across tests and different chaotic training signals. Although analytical
studies are limited, we introduce a simple ‘toy model’ to qualitatively describe some
properties. We conclude our work with study of the statistics of trained output weights,
which also exhibit universal characteristics. These universalities provide deeper insights
into the inner workings and behaviour of ESNs, enhancing our understanding of how
information spreads throughout the network.EPSRC DPT PhD studentship schem
Connected Directors–Advisors and Mergers and Acquisitions Outcomes
Get PDFSupporting Information is available online at: https://onlinelibrary.wiley.com/doi/10.1111/1467-8551.12860#support-information-section .[Correction added on 2 April 2025, after first online publication: The affiliation for the author “Yizhe Dong” has been updated in this version.]We examine the impact of social ties between directors (directors or senior managers) of acquiring firms and their corresponding advisors on several mergers and acquisitions outcomes. The social ties between acquirers’ directors and their advisors are positively related to acquirers’ gains in the short and long run. This is due to lower takeover premia, lower advisory fees and shorter period to deal completion. This relation is more pronounced in deals involving inexperienced acquirers, targets in more opaque industries and targets recommended by advisors. We also find that acquirers are more likely to withdraw from deals with high premia if they hire a socially connected advisor. Identification is addressed by using instrumental variables, excluding deals that are prone to endogeneity and using the propensity score matching and the entropy balancing methods
Threshold Sensitivity in Two-Channel Modulo ADCs: Analysis and Robust Reconstruction
Get PDFThis paper presents a comprehensive analysis of two-channel modulo analog-to-digital converters (ADCs) systems, focusing on the sensitivity of ADC thresholds. By exploiting analytic number theory, we first investigate the relationship among ADC threshold precision, maximum signal dynamic range, and error tolerance. Our analysis reveals that even slight deviations in ADC thresholds can substantially impact the maximum reconstructed signal dynamic range and error tolerance. To address these sensitivity issues, we propose a novel approach that strategically sacrifices signal dynamic range to stabilise error tolerance in the presence of slight ADC threshold variations. We also introduce a low-complexity reconstruction algorithm that exploits this trade-off, thereby enhancing system robustness. Simulation results validate the theoretical framework and confirm the efficiency of our proposed algorithm
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! 👇