299645 research outputs found
Sort by
A multi-agent reinforcement learning (MARL) framework for designing an optimal state-specific hybrid maintenance policy for a series k-out-of-n load-sharing system
No full textThe series k-out-of-n: G load-sharing structure is widely adopted in engineering. During their operations, system components are subject to deterioration that causes system failures and shutdowns. Although maintenance reduces system failure-associated costs, it also requires system shutdown and incurs considerable costs. This calls upon a maintenance policy that minimizes the overall long-term cost rate. When the components have continuous and load-dependent deterioration processes and the maintenance duration is non-negligible, the task becomes especially challenging. In this paper, we propose a Markov decision process (MDP)-based multi-agent reinforcement learning (MARL) framework to obtain an optimal state-specific hybrid maintenance policy that determines the maintenance timing and levels for all components holistically. First, we define the policy that dictates whether each component undergoes imperfect repair or replacement at periodic decision epochs. Second, we establish an MDP-based multi-agent framework to quantify the system's cost rate by defining the state and action spaces, modeling the stochastic transitions of components’ dependent deterioration processes, and formulating a well-calibrated penalty function. Third, we customize a MARL algorithm which leverages neural networks to handle the large state space and integrates the Branching Dueling Network structure to decompose the high-dimensional action space, thereby improving the scalability. A heuristic-enhanced penalty function is designed to avoid suboptimal policies. A power plant case study demonstrates the effectiveness of the proposed policy and underscores the importance of accounting for maintenance duration in policy design
The Role of Chinese Arbitration Centres in Environmental, Social and Governance (ESG) Disputes
No full textMediation Programme Origins and Practice from a Chinese and Multi-Jurisdictional Perspective
No full textMIB-Net: Balance the mutual information flow in deep learning network for multi-dimensional segmentation of COVID-19 CT images
No full textComputed tomography (CT) based lung screening followed by computer-vision-based automated segmentation has great potential in early diagnosis of COVID-19, but low sample size and high patient-to-patient variability of infection characteristics are two of the major challenges. We propose a novel Mutual Information Balanced Net (MIB-Net) with three distinctive features. First, it captures mutual information across feature maps, CT scans, and predicted segmentation labels to optimize the contribution of different feature maps for better segmentation accuracy. Second, it incorporates a three-dimensional structural prior of CT scans based on predicted segmentation of neighboring CT slices to improve accuracy. Third, a novel Weight Optimization Filter (WOF) is used to adjust the proportion of 3D structural prior to be incorporated, which further improves the segmentation accuracy. Experimental results show that the proposed approach outperforms state-of-the-art biomedical image segmentation models by 7.87%, 3.85%, and 5.01% in Dice Score, respectively for three popular COVID-19 CT datasets. The better performance of the proposed approach suggests that it may serve as an attractive alternative to Lung CT-based screening for COVID-19.</p
Molecular Engineering of N-heteroaromatic Organic Cathode for High-Voltage and Highly Stable Zinc Batteries
No full textZinc batteries hold promise for grid-scale energy storage due to their safety and low cost. A key challenge for the field is identifying cathode materials that can undergo reversible redox reactions at the extreme potentials required for realizing high energy density devices. While organic materials have been extensively explored as cathode materials due to their structural tunability and eco-friendliness, most reported zinc-organic batteries exhibit a voltage lower than 1.2 V. In this report, by employing rational molecular design and synthesis, computational analysis, and electrochemical evaluation, the well-studied neutral p-type N-centered is redesigned, triphenylamine organic cathode by replacing three phenyl rings with the smallest aromatic system – cationic cyclopropenium. This results in a novel class of cathode materials with simultaneously enhanced potential, capacity, and stability. The resultant full battery exhibits a high discharge voltage of 1.7 V and an outstanding capacity retention of 95% after 10000 cycles at a discharge capacity of 157.5 mAh g−1cation (103.9 mAh g−1salt).link_to_subscribed_fulltex
Modulating the Leverage Relationship in Nitrogen Fixation Through Hydrogen-Bond-Regulated Proton Transfer
No full textIn the electrochemical nitrogen reduction reaction (NRR), a leverage relationship exists between NH3-producing activity and selectivity because of the competing hydrogen evolution reaction (HER), which means that high activity with strong protons adsorption causes low product selectivity. Herein, we design a novel metal-organic hydrogen bonding framework (MOHBF) material to modulate this leverage relationship by a hydrogen-bond-regulated proton transfer pathway. The MOHBF material was composited with reduced graphene oxide (rGO) to form a Ni-N2O2 molecular catalyst (Ni-N2O2/rGO). The unique structure of O atoms in Ni-O-C and N-O-H could form hydrogen bonds with H2O molecules to interfere with protons being directly adsorbed onto Ni active sites, thus regulating the proton transfer mechanism and slowing the HER kinetics, thereby modulating the leverage relationship. Moreover, this catalyst has abundant Ni-single-atom sites enriched with Ni-N/O coordination, conducive to the adsorption and activation of N2. The Ni-N2O2/rGO exhibits simultaneously enhanced activity and selectivity of NH3 production with a maximum NH3 yield rate of 209.7 μg h−1 mgcat.−1 and a Faradaic efficiency of 45.7 %, outperforming other reported single-atom NRR catalysts.link_to_subscribed_fulltex
Effective Water Confinement and Dual Electrolyte–Electrode Interfaces by Zwitterionic Oligomer for High-Voltage Aqueous Lithium-Ion Batteries
No full textAqueous lithium-ion batteries (ALIBs) have attracted significant interest due to their inherent advantage on safety. However, water itself has a narrow electrochemical stability window (ESW), limiting the energy density of ALIBs. Here, a low-molecular-weight zwitterionic oligomer, oligo(propylsulfonate dimethylammonium propylmethacrylamide) (OPDP), as an effective water binding agent for high-voltage ALIBs is demonstrated. The OPDP can effectively confine water molecules while reducing water activity. The OPDP-based electrolyte, with an ultra-high water weight percentage of 25.4%, possesses an outstanding ESW of up to 3.26 V and an ionic conductivity as high as 3.18 mS cm−1. Furthermore, the aqueous Mo6S8//LiMn2O4 full cell with OPDP-based electrolyte achieves a 99.7% capacity retention after 200 cycles at 0.5C with a high Coulombic efficiency (CE) of 98.7% and a specific energy of 88–101 Wh kg−1. Also, it achieves an 89% capacity retention after 2000 cycles at 10C with a high CE of 99.9%. These postmortem characterizations suggest that robust organic–inorganic hybrid cathode/anode-electrolyte interfaces have been constructed during the cycling through the heteroatoms of N, S, and O in the zwitterionic oligomer, leading to the inhibited hydrogen/oxygen evolution reactions and high performance of the full cell. This work provides a promising strategy for developing low-cost and high-voltage aqueous batteries.link_to_subscribed_fulltex
Enhancing dialogic teaching in Hong Kong ESL classrooms through video based professional development
No full textThis dissertation investigates the design, implementation, and evaluation of a video-based teacher professional development (PD) programme for developing dialogic teaching in Hong Kong English as a Second Language (ESL) classrooms. Dialogic teaching, among its many features, highlights teachers’ skillful extension of student responses to encourage elaboration and reasoning, facilitating authentic exchanges of meaning.
Specifically, this work aimed to 1) explore trends in video-based teacher PD through bibliometric analysis; 2) examine the link between dialogic teaching, student discursive engagement, and student second language (L2) learning; 3) investigate the effectiveness of a video-based PD programme as a whole and its elements in enhancing dialogic teaching; and 4) reveal the teacher learning process during collaborative coding in the programme.
These aims were addressed through four interrelated studies. Video-based PD trends were examined in Study I through analysing the Scopus database, providing insights for developing the PD programme for Studies III and IV, especially on the concept of teaching noticing. Studies II to IV involved 27 teachers and their students from seven Hong Kong primary and secondary schools, with the teachers participating in a six-month video-based PD programme under three different dose conditions in Studies III and IV. In Study II, structural equation modelling (SEM) analysed student questionnaires (n = 490) to explore the relationships among teacher dialogic teaching, student discursive engagement, and L2 learning. Study III assessed the impact of dialogic teaching and PD elements by comparing pre- and post-intervention and the three doses, using data from student questionnaires (n = 400), lesson video recordings, and teacher interviews. Study IV further revealed the teacher learning process by examining interaction in collaborative coding workshops.
The research yielded five major findings: (1) the literature indicated that video-based PD programmes emphasised the theme of ‘noticing’, yet there was a notable gap researching humanities subjects, despite increasing interest in classroom discourse analysis; (2) extending prior research on L2-related characteristics including motivation, communication competence, and willingness to communicate (WTC), the SEM indicated that WTC in class was a mediator between students’ perceptions of teachers using productive classroom talk and their discursive engagement with other students; (3) a pre-post comparison indicated that implementing dialogic teaching enhanced ESL classroom efficacy, as evidenced by student questionnaires, teacher video recordings, and interviews; (4) a dose-response analysis showed that a three-dose intervention of video-based PD was more effective than two-dose and one-dose conditions to some extent, highlighting the importance of diverse PD elements and duration for effective teacher inquiry; and (5) collaborative coding facilitated the development of noticing in teachers for dialogic teaching.
This dissertation examines how individual student characteristics, such as motivation, and teachers’ mindful use of dialogic teaching influence student classroom participation. It also identifies crucial components of effective PD and illustrates how collaborative video coding supports the practice of teacher noticing. From a pedagogical standpoint, it demonstrates that integrating dialogic teaching through a video-based PD programme enhances ESL classroom effectiveness, providing valuable insights for PD designers and educators to improve L2 teaching and learning.published_or_final_versionEducationDoctoralDoctor of Philosoph
Innovation for social value : three essays on pay equity, healthcare delivery, and drug development
No full textIn recent years, research on social value creation has gained increasing attention, with a growing body of literature investigating how organizations can contribute to addressing social needs and challenges. Scholars have drawn upon various theoretical perspectives, such as social learning theory, stakeholder theory, and institutional theory, to explain the antecedents and processes of social value creation. To add to this stream of literature, I conduct three studies to investigate the role of innovation in enabling organizations to generate social value and to examine how the institutional environment influences innovation that contributes to social value creation.
The first essay examines the influence of AI orientation on human value within organizations. Given the emergence of AI, concerns arise about its potential to diminish the value and contributions of human labor, which can lead to severe ethical problems, such as the deterioration of employee well-being, increased job stress, and a sense of deprivation. This study examines how AI orientation affects human value within organizations by investigating its impact on a firm’s labor income share and top management team (TMT)-employee pay gap. By analyzing a sample of Chinese-listed firms from 2015 to 2022, we find that a firm’s AI orientation is positively related to its labor income share and negatively related to the TMT-employee pay gap. We also show that these effects are stronger for knowledge-intensive firms, and the negative effect of AI orientation on the pay gap is weaker when firms’ TMT tenure is long. These findings contribute to the business ethics and AI literatures by revealing the beneficial and contingent roles of AI in preserving human value.
The second essay investigates the role of intelligent support systems (ISS) in moderating fatigue’s impact on healthcare delivery. Focusing on stroke care, we analyze the field trial of StrokePro, an AI-driven ISS designed to improve care delivery. Using a difference-in-differences approach with coarsened exact matching on a stroke care clinical records dataset, we assess the role of ISS in moderating fatigue’s impact on caregiver guideline adherence. We found that ISS helps alleviate the negative effects of fatigue on caregiver guideline adherence, particularly for patients with high disease severity and complex diagnoses, and for less proficient caregivers.
The third essay investigates how process regulation influences product innovation. While existing research has widely examined how regulations affect innovation inputs and outputs, we know little about the role of regulation that targets the innovation process. We argue that process regulation may decrease firms’ product innovation because it slows the development process and heightens pipeline abandonment chances. Leveraging a clinical trial inspection enforced by the China Food and Drug Administration as an exogenous shock, we assess the impact of process regulation on product innovation in the pharmaceutical industry. Employing a difference-in-differences approach, we find that clinical trial inspection reduces firms’ product innovation. We further show that the negative effect of clinical trial inspection on product innovation is weaker for firms with high R&D intensity or foreign shareholding. These findings extend our understanding of regulation’s influence on innovation and provide important implications for pharmaceutical firms and policymakers.published_or_final_versionBusinessDoctoralDoctor of Philosoph
Advanced numerical simulation on effects of welding onto high strength S690 and S960 steel welded H-sections
No full textIn this paper, an advanced numerical simulation approach, namely, the “Thermal-Metallurgical-Mechanical-Structural” approach, or the TMMS approach, is presented which is able to simulate the effects of welding on the structural behaviour of high strength S690 and S960 welded sections considering phase transformations. This approach is an extension of the “Integrated and Coordinated Advanced Numerical Simulation” approach, or the ICANS approach, which was also proposed by the authors. It should be noted that no phase transformation is considered in the ICANS approach. The ICANS approach is first presented in this paper with a thorough description on both thermal and thermo-mechanical responses of these high strength steel using Abaqus under the effects of welding. The TMMS approach is then presented with an emphasis on thermo-metallurgical responses of the high strength steel considering phase transformations using SYSWELD under the effects of welding. Various sets of measured data on thermal, thermo-metallurgical, thermo-mechanical and structural responses of a total of four welded H-sections are presented, and detailed comparison between measured and numerical data is presented. Owing to occurrence of phase transformations, new material types with different mechanical properties were formed within the heat affected zones, and their phase-dependent mechanical properties as well as residual stresses and plastic strains were incorporated into structural models for structural analysis. The predicted deformation characteristics of the models were demonstrated to follow those of the measured data of welded H-sections under compression closely along the entire deformation ranges