Zheng, RuiyanNg, S. ThomasShao, YuyangLi, Zhongfu...
1.1-1.18页
查看更多>>摘要：In the post COVID-19 era, there is an escalating demand to fundamentally rethink and digitalize healthcare emergency management (HEM) to ensure greater resilience and responsiveness. Among emerging technologies, the digital twin (DT) holds unique promise by enabling real-time monitoring, dynamic decision support, and predictive maintenance, all of which are critical in high-stakes emergency scenarios. Despite its potential, DT deployment in HEM remains an intricate, long-term endeavor, hampered by significant conceptual and technical barriers. Many stakeholders lack a clear understanding of DT's functional scope, the requisite technologies for robust implementation, and pathways for integrating DT into established healthcare workflows. In response, this paper offers a comprehensive examination of DT in HEM, categorizing current applications across four levels: individual, hospital, public, and cloud supporting. This paper also highlights how contemporary technical solutions, ranging from advanced networking and distributed computing to AI-driven analytics, can be orchestrated to support novel DT functionalities in real-world healthcare operations. Additionally, challenges, open problems and future directions for DT in HEM are discussed. By synthesizing both functional and research-oriented insights, this review aims to clarify future directions for leveraging DT as a transformative vehicle for healthcare emergency preparedness, response, and long-term resilience.
查看更多>>摘要：Human reliability analysis (HRA) plays a pivotal role in safety-critical systems, with its methodological evolution currently advancing into the third generation, characterized by dynamic modeling and deeper cognitive processing frameworks. In this study, we propose a novel paradigm extension to HRA, introduced within an emergent operational environment. Specifically, we develop a dynamic risk-informed framework (DRIF) that integrates Bayesian networks (BNs), long short-term memory (LSTM) neural networks, and domain-specific emergency operating procedures (EOPs) to enable real-time evaluation of human error risks during emergency scenarios. The framework employs Bayesian networks to probabilistically model causal relationships among human factors, while LSTM networks dynamically process temporal operational data streams for fault diagnosis. This hybrid architecture synergizes HRA principles with real-time risk propagation mechanisms, thereby enhancing situational awareness and decision granularity under time-critical conditions. To empirically validate DRIF's efficacy, we implemented it in anomaly mission scenarios for a high-temperature gas-cooled reactor (HTGR). The case study demonstrates the framework's capability to (1) quantify human error probabilities (HEPs) through probabilistic inference, (2) identify latent risk pathways via backward propagation analysis, and (3) provide prescriptive guidance aligned with EOPs for risk mitigation. The results show that the more precisely later emergency action measures are implemented, the better the accident prevention and control effect during emergencies. This advancement establishes a methodological foundation for next-generation HRA systems in complex engineered systems.
查看更多>>摘要：Falling accidents have a high probability of occurring during the construction of mega hydropower engineering (MHE), involving various risk factors. Previous studies have paid limited attention to the decision-making management of risk pre-control measures. To address this, we propose a text data-driven decision analysis method. First, we couple fault tree analysis (FTA) with grey relational analysis (GRA), developing the FTA-GRA method to assess the importance of risk factors, identify main factors, and formulate pre-control measures. Subsequently, the Cloud Model (C) and TODIM (an acronym in Portuguese for Interactive and Multi-criteria Decision Making) were used to improve Multi-Attributive Border Approximation Area Comparison (MABAC), and the C-TODIM-MABAC decision-making method was constructed to quantify the priority of each pre-control measure to select the optimal solution. The results show that (1) among the top 10 risk factors in terms of importance, workers' non-compliant operations and neglect of safety issues are particularly prominent. (2) The C-TODIM-MABAC method fully considers the fuzziness and randomness of evaluation language and the bounded rationality of decision-makers, making it more suitable for MHE construction scenarios. (3) Sensitivity analysis verifies the applicability and reliability of the method from the dual perspectives of criterion weights and evaluation scales.
查看更多>>摘要：This paper presents a probabilistic seismic demand model (PSDM) as a relationship between intensity measures (IMs) and engineering demand parameters (EDPs) for the seismic assessment of two case studies resembling historical masonry buildings. The first one is representative of stiff monumental buildings, and the second of tall and slender masonry buildings. Both structures are modelled in the OpenSees software using three-dimensional macroelements that consider both the in-plane and out-of-plane response of masonry walls. A set of 100 accelerograms are selected to represent the seismic excitation. After full characterization of the seismic input in terms of IMs, both buildings are subjected to the action of these accelerograms to study the maximum structural response in the context of cloud analysis. The most suitable IMs are determined subsequently under the notions of efficiency, practicability, proficiency, and sufficiency. In addition, a composed measure is proposed as a linear combination in logarithmic space of the IMs that exhibit the best coefficient of determination (R2) within the EDP vs. IM regression. This optimal composed measure is determined through machine learning-based Lasso regression. In the final stage of the study, fragility curves are derived to measure the likelihood of exceedance of certain levels of average roof displacement in terms of IM parameters.
查看更多>>摘要：Accurate Remaining Useful Life (RUL) prediction model relies on full-lifecycle degradation features of the equipment. However, fragmented out-of-distribution (OOD) data due to specific working condition, equipment service time and communication packet loss inevitably affect the prediction accuracy. This study proposes a personalized federated RUL prediction method for fragmented OOD data scenarios, aiming to integrate OOD data fragments provided by different clients. In this means, a federated prediction model can be established to capture the full-lifecycle degradation features by incorporating fragmented OOD data. We focus on establishing a correctable cycle-consistent alignment mechanism driven by health state similarity to solve the challenging problem arisen by inter-client spatiotemporal heterogeneity. A novel health assessment index based on the quantile of hypothesis test is designed to capture the degradation feature required in the cycle-consistent alignment mechanism. Once new fragmented OOD data is available, a personalized federation strategy is developed by designing an adversarial mechanism between degradation features involved in the previous old OOD data and the new OOD data, such that previous degradation features can be further extended to a more full degradation feature. The superiority of the proposed method in RUL prediction was validated on fragmented OOD data collected on benchmark bearing prognostic system (BPS) platform.
查看更多>>摘要：The wear modeling and life prediction of pantographs are crucial for ensuring the safety and reliability of urban rail transit systems. However, because of the complex interplay between stochastic vibrations and electrical currents, pantograph wear exhibits strong variability, and physics-based degradation predictions based solely on material parameters, environmental factors, and wear mechanisms are limited in accuracy. Purely data-driven approaches, on the other hand, are constrained by their reliance on large datasets and lack of interpretability, making them difficult to meet practical engineering needs. To address these challenges, we propose an interpretable variational model called IV-NBEATS. This study integrates the surface wear mechanism under the asperity hypothesis into the N-BEATS model using the projection principle, thereby enhancing the interpretability of the model. In addition, we introduce a method for describing the uncertainty of key wear parameters, enabling a deep network to represent the uncertainty of these parameters. Furthermore, to cope with dynamic changes in wear system parameters, we propose a dynamic updating method based on a Bayesian directed graph model that effectively overcomes the limitations of existing methods in capturing the temporal evolution of wear system parameters. Finally, the effectiveness of the proposed approach is demonstrated through the analysis of a real-world case study of pantograph wear.
查看更多>>摘要：The digital twin method is a foundational technology for the digitization and intelligence of complex nuclear power equipment, e.g. the control rod drive mechanism. There is an urgent need to develop digital twin modeling methods and application platforms tailored for the safety of complex technical systems. However, current digital twin modeling techniques struggle to meet the requirements for real-time fault monitoring and long-term predictive maintenance simultaneously. Therefore, a five-dimensional digital twin modeling architecture based on a dual system for safety, which combines an offline digital twin system for reliability and fatigue analysis with an online digital twin system for real-time fault monitoring, has been proposed in the current study. The control rod drive mechanism digital twin platform, developed in Isight, is designed to incorporate a dual system for both real-time monitoring and long-term predictive maintenance. A deviation of less than 5% is maintained between operational and experimental data, thus enhancing the reliability and performance of control rod drive mechanism system.
查看更多>>摘要：Intelligent agents often encounter challenges in balancing safety and performance when transitioning from general training scenarios to specific task scenarios due to unknown environmental differences. Under the uncertainty of new scenarios, safety considerations constrain extensive exploration, resulting in limited policy improvement. This paper proposes a novel reinforcement learning approach featuring a dependable policy improvement algorithm that emphasizes safety and confidence throughout the entire training process. The proposed algorithm enhances the baseline policy developed in general training scenarios to guide exploration and designs confidence bounds to evaluate both task performance and safety. By cautiously exploring and updating policies based on data confidence bounds, the approach ensures reliable agent behavior in new, uncertain, and potentially risky environments. Simulation experiments with an automatic guided vehicle (AGV) demonstrate the effectiveness of this approach across various scenarios.
查看更多>>摘要：This study explores how to maintain the continuous operation of infrastructure in the face of disruptions. We demonstrate that a connection exists between the established fields of cybernetics and resilience, and that the resilience of a system can be assessed and enhanced through the application of methods from cybernetics and control theory. By adding a closed-loop feedback controller that adapts the investment in maintenance according to the current conditions, those in charge of infrastructure can be ready for problems, adapt to them, and bounce back quickly. The novelty of this work is not only acknowledging the delay between measuring a problem and implementing a solution but also in addressing the implications of this delay when configuring the controller and bringing this knowledge into the field of resilience. We stress the idea that working together across different fields can help us deal with problems more effectively. We apply this concept to a system-theoretic infrastructure model to present how this could work in practice. By figuring out these solutions, we hope to make infrastructure more resilient and ready for whatever comes its way.
查看更多>>摘要：Intelligent fault diagnosis technology determines the safety and reliability of equipment operation, and domain-based adaptive fault diagnosis models have been explored for solving the problem of data distribution discrepancies caused by different operating conditions. However, the requirement of obtaining the unlabeled target domain data in advance limits its application in real-world equipment operating scenarios. To address this problem, this paper proposes an inter-domain multivariate linearization (IML)-guided domain generalization network (IMLNet) for intelligent fault diagnosis. A domain multivariate fusion generation module is designed to construct new domains by linearizing between different domains using inter-domain multivariate linearization, which helps the network to extract domain invariant features in depth. Meanwhile, by fusing the multi-attention mechanism and feature pyramid network on the basis of residual network, it promotes the network to capture multi-scale information and provide richer semantic information. The effectiveness of the method is verified through two different fault diagnosis cases.
NETL
NSTL
Elsevier