Search Results (10,520)

This paper introduces an iceberg-based digital risk twin (DRT) framework for the health monitoring of complex engineering systems. The proposed model transforms multidimensional sensor and contextual data into a structured, interpretable three-dimensional geometry that captures both observable and latent risk components. Each monitored parameter is represented as a vertical geometric sheet whose height encodes a normalized risk level, producing an evolving iceberg structure in which the visible and submerged regions distinguish emergent anomalies from latent degradation. A formal mathematical formulation is developed, defining the mappings from the risk vector to geometric height functions, spatial layout, and surface composition. The resulting parametric representation provides both analytical tractability and intuitive visualization. A case study involving an aircraft fuel system demonstrates the capacity of the DRT to reveal dominant risk drivers, parameter asymmetries, and temporal trends not easily observable in traditional time-series analysis. The model is shown to integrate naturally into AI-enabled health management pipelines, providing an interpretable intermediary layer between raw data streams and advanced diagnostic or predictive algorithms. Owing to its modular structure and domain-agnostic formulation, the DRT approach is applicable beyond aviation, including power grids, rail systems, and industrial equipment monitoring. The results indicate that the iceberg representation offers a promising foundation for enhancing explainability, situational awareness, and decision support in the monitoring of complex engineering systems. Full article

This study is essential for medium- and long-term land-use management, as land-use patterns directly influence local economic and social development. Geographic Information System (GIS) techniques are fundamental tools for analyzing a wide range of geomorphological processes, including relief fragmentation density, relief energy, soil texture, slope gradient, and slope orientation. The present research focuses on the Pesceana river basin in the Southern Carpathians, Romania. It addresses three main objectives: (1) to analyze land-use dynamics derived from CORINE Land Cover (CLC) data between 1990 and 2018, along with the long-term distribution of the Normalized Difference Vegetation Index (NDVI) for the period 2000–2025; (2) to evaluate the basin’s natural potential byintegrating topographic data (contour lines and profiles) with relief fragmentation density, relief energy, vegetation cover, soil texture, slope gradient, aspect, the Stream Power Index (SPI), and the Topographic Wetness Index (TWI); and (3) to assess the spatial distribution of habitat types, characteristic plant associations, and soil properties obtained through field investigations. For the first two research objectives, ArcGIS v. 10.7.2 served as the main tool for geospatial processing. For the third, field data were essential for geolocating soil samples and defining vegetation types across the entire 247 km² area. The spatiotemporal analysis from 1990 to 2018 reveals a landscape in which deciduous forests clearly dominate; they expanded from an initial area of 80 km² in 1990 to over 90 km² in 2012–2018. This increase, together with agricultural expansion, is reflected in the NDVI values after 2000, which show a sharp increase in vegetation density. Interestingly, other categories—such as water bodies, natural grasslands, and industrial areas—barely changed, each consistently representing less than 1 km² throughout the study period. These findings emphasize the importance of land-use/land-cover (LULC) data within the applied GIS model, which enhances the spatial characterization of geomorphological processes—such as vegetation distribution, soil texture, slope morphology, and relief fragmentation density. This integration allows a realistic assessment of the physical–geographic, landscape, and pedological conditions of the river basin. Full article

The global clean energy transition requires power conversion technologies that combine high efficiency, operational flexibility, and reduced environmental impact over their entire service life. Solid-state transformers (SSTs) have emerged as a promising alternative to conventional line-frequency transformers, offering bidirectional power flow, high-frequency isolation, and advanced control capabilities that support renewable integration and electrified infrastructures. This paper presents a comparative life cycle assessment (LCA) of conventional transformers and SSTs across representative power-system applications, including residential and industrial distribution networks, electric vehicle fast-charging infrastructure, and transmission–distribution interface substations. The analysis follows a cradle-to-grave approach and is based on literature-derived LCA data, manufacturer specifications, and harmonized engineering assumptions applied consistently across all case studies. The results show that, under identical assumptions, SST-based solutions are associated with indicative lifecycle CO₂ emission reductions of approximately 10–30% compared to conventional transformers, depending on power rating and operating profile (≈90–1000 t CO₂ over 25 years across the four cases). These reductions are primarily driven by lower operational losses and reduced material intensity, while additional system-level benefits arise from enhanced controllability and compatibility with renewable-rich and hybrid AC/DC grids. The study also identifies key challenges that influence the sustainability performance of SSTs, including higher capital cost, thermal management requirements, and the long-term reliability of power-electronic components. Overall, the results indicate that SSTs represent a relevant enabling technology for future low-carbon power systems, while highlighting the importance of transparent assumptions and lifecycle-oriented evaluation when comparing emerging grid technologies. Full article

The increasing complexity of modern electrical networks poses significant challenges in terms of monitoring, maintenance, and operational efficiency. However, current planning approaches often lack a unified integration of real-time data and predictive modeling. In this context, Digital Twins (DTs) emerge as a promising solution, as they enable the creation of virtual replicas of physical assets. This research addresses the lack of standardized technical frameworks by proposing a novel mathematical optimization model for grid planning based on DTs. The proposed methodology integrates comprehensive architecture (frontend/backend), specific data standards (IEC 61850), and a linear optimization formulation to minimize operational costs and enhance reliability. Case studies such as DTEK Grids and American Electric Power are analyzed to validate the approach. The results demonstrate that the proposed framework can reduce planning errors by approximately 15% and improve fault prediction accuracy to 99%, validating the DTs as a key tool for the digital transformation of the energy sector towards Industry 5.0. Full article

PM_2.5 pollution has become one of the major environmental issues in Shandong Province in recent years. High concentrations of PM_2.5 not only reduce atmospheric visibility but also induce respiratory and cardiovascular diseases, and significantly increase health risks. Source apportionment of PM_2.5 is important for policy makers to determine control strategies. This study analyzed regional and sectoral PM_2.5 sources across 17 Shandong cities during the 2017–2018 winter heating season, which is selected because it is representative of severe air pollution with an average PM_2.5 of 65.75 μg/m³ and hourly peak exceeding 250 μg/m³. This air pollution episode aligned with key control policies, where seven major cities implemented steel capacity reduction and coal-to-gas/electric heating, as a baseline for evaluating emission reduction effectiveness. The particulate matter source apportionment technology in the Comprehensive Air Quality Model with extensions (CAMx) was applied to simulate the source contributions to PM_2.5 in 17 cities from different regions and sectors including industry, residence, transportation, and coal-burning power plants. The meteorological fields required for the CAMx model were generated using the Weather Research and Forecasting (WRF) model. The results showed that all cities besides Dezhou city in Shandong Province contributed PM_2.5 locally, varying from 39% to 53%. The emissions from Hebei province have a large impact on the PM_2.5 concentrations in Shandong Province. The non-local industrial and residential sources in Shandong Province accounted for the prominent proportion of local PM_2.5 in all cities. The contribution of non-local industrial sources to PM_2.5 in Heze City was up to 56.99%. As for Zibo City, the largest contribution of PM_2.5 was from non-local residential sources, around 56%. Additionally, the local industrial and residential sources in Jinan and Rizhao cities had relatively more contributions to the local PM_2.5 concentrations compared to the other cities in Shandong Province. Finally, the emission reduction effects were evaluated by applying different reduction ratios of local industrial and transportation sources, with decreases in PM_2.5 concentrations ranging from 0.2 to 26 µg/m³ in each city. Full article

This paper proposes a PI-Dual-STGCN fault diagnosis model based on a SHAP-LLM joint explanation framework to address issues such as the lack of transparency in the diagnostic process of deep learning models and the weak interpretability of diagnostic results. PI-Dual-STGCN enhances the interpretability of graph data by introducing physical constraints and constructs a dual-graph architecture based on physical topology graphs and signal similarity graphs. The experimental results show that the dual-graph complementary architecture enhances diagnostic accuracy to 99.22%. Second, a general-purpose SHAP-LLM explanation framework is designed: Explainable AI (XAI) technology is used to analyze the decision logic of the diagnostic model and generate visual explanations, establishing a hierarchical knowledge base that includes performance metrics, explanation reliability, and fault experience. Retrieval-Augmented Generation (RAG) technology is innovatively combined to integrate model performance and Shapley Additive Explanations (SHAP) reliability assessment through the main report prompt, while the sub-report prompt enables detailed fault analysis and repair decision generation. Finally, experiments demonstrate that this approach avoids the uncertainty of directly using large models for fault diagnosis: we delegate all fault diagnosis tasks and core explainability tasks to more mature deep learning algorithms and XAI technology and only leverage the powerful textual reasoning capabilities of large models to process pre-quantified, fact-based textual information (e.g., model performance metrics, SHAP explanation results). This method enhances diagnostic transparency through XAI-generated visual and quantitative explanations of model decision logic while reducing the risk of large model hallucinations by restricting large models to reasoning over grounded, fact-based textual content rather than direct fault diagnosis, providing verifiable intelligent decision support for industrial fault diagnosis. Full article

This study proposes a neural-augmented Libor Market Model (LMM) for swaption surface calibration that enhances expressive power while maintaining the interpretability, arbitrage-free structure, and numerical stability of the classical framework. Classical LMM parametrizations, based on exponential decay volatility functions and static correlation kernels, are known to perform poorly in sparsely quoted and long-tenor regions of swaption volatility cubes. Machine learning–based diffusion models offer flexibility but often lack transparency, stability, and measure-consistent dynamics. To reconcile these requirements, the present approach embeds a compact neural network within the volatility and correlation layers of the LMM, constrained by structural diagnostics, low-rank correlation construction, and HJM-consistent drift. Empirical tests across major currencies (EUR, GBP, USD) and multiple quarterly datasets from 2024 to 2025 show that the neural-augmented LMM consistently outperforms the classical model. Improvements of approximately 7–10% in implied volatility RMSE and 10–15% in PV RMSE are observed across all datasets, with no deterioration in any region of the surface. These results reflect the model’s ability to represent cross-tenor dependencies and surface curvature beyond the reach of classical parametrizations, while remaining economically interpretable and numerically tractable. The findings support hybrid model designs in quantitative finance, where small neural components complement robust analytical structures. The approach aligns with ongoing industry efforts to integrate machine learning into regulatory-compliant pricing models and provides a pathway for future generative LMM variants that retain an arbitrage-free diffusion structure while learning data-driven volatility geometry. Full article

Carbon capture is an essential technology for reducing industrial CO₂ emissions, particularly in the power and cement sectors. Among the various capture methods, solvent-based absorption systems are widely used due to their efficiency and scalability, making the selection of the right solvent critical for near-term applications. This study analyzes several solvents for use in an absorption-based CO₂ capture system, emphasizing identifying the most suitable solvent for 2025–2030. The research methodology involves process modeling in Aspen Plus, sensitivity analysis, and evaluation of the regeneration duty for each solvent. The objective is to achieve at least 90% CO₂ capture and 95% CO₂ purity. The flue gas composition considered in this analysis is 19.8% CO₂, 9.3% O₂, 63% N₂, 7.5% H₂O, and other trace gases. Various solvents are evaluated to determine their effectiveness in capturing CO₂ while minimizing the energy consumption during solvent regeneration. A sensitivity analysis was conducted to optimize the system’s performance based on the solvent type, operating conditions, and regeneration duty. The results showed that amine blends demonstrated a CO₂ capture rate of 92% and a CO₂ purity of 96%, with regeneration energy requirements of around 3.2 GJ/ton of CO₂, significantly lower than those of traditional MEA systems, which typically require around 4.0 GJ/ton. In contrast, ionic liquids showed a CO₂ capture rate of 89% and a purity of 95%, with a regeneration energy of 2.8 GJ/ton, though their current cost is higher, limiting their immediate large-scale application. Annual capital expenditure (CAPEX) calculation revealed that amine blends could potentially reduce the CAPEX by 15–20% compared to MEA, while amino acid salts showed similar CAPEX reductions with a capture efficiency of 90%. Overall, the results indicate that hybrid amine solvents are the most cost-effective and practical solution for 2025–2030, with ionic liquids and amino acid salts emerging as promising alternatives as their costs decrease. Full article

Modular construction (MC) offers significant environmental and efficiency advantages yet maintains low market penetration in China despite substantial government support. This study addresses the critical knowledge gap by systematically analyzing complex barrier interrelationships across project phases and stakeholder groups (university, construction authority, supplier/manufacturer company) to develop a comprehensive MC promotion framework. A four-phase mixed method approach was employed. (1) Grounded theory analysis of MC policy frameworks was performed in Singapore, the United States, and Hong Kong to extract best practice insights. (2) A systematic literature review and multi-round Delphi expert consultations were used to identify 21 core barriers across six project stages (decision-making, procurement, design, production, transportation, and construction acceptance). (3) The DEMATEL analysis reveals causal relationships among barriers based on experts’ perceived influence between factors. (4) Integrated ISM-MICMAC methodology was used to establish hierarchical structures and barrier classifications. Institutional barriers emerged as the primary impediment to MC diffusion, with unclear authority distribution between government administrations and design organizations identified as the most critical factor. The MICMAC analysis categorized the 21 barriers into four distinct groups based on their driving power and dependence characteristics, revealing complex causal relationships among barriers across the six project stages while highlighting the emergent role of higher education institutions in industrial transformation. Successful MC implementation requires market-oriented, context-specific strategies prioritizing institutional framework development, with the findings providing actionable insights for policymakers to address regulatory ambiguities and practical guidance for industry practitioners developing targeted MC promotion strategies in emerging markets. Full article

Massive MIMO systems can substantially enhance spectral efficiency, but such gains rely on the availability of accurate channel state information (CSI). However, the increase in the number of antennas leads to a significant growth in feedback overhead, while conventional deep-learning-based CSI feedback methods also impose a substantial computational burden on the user equipment (UE). To address these challenges, this paper proposes LLM4FB, a one-sided CSI feedback framework that leverages a pre-trained large language model (LLM). In this framework, the UE performs only low-complexity linear projections to compress CSI. In contrast, the BS leverages a pre-trained LLM to accurately reconstruct and predict CSI. By utilizing the powerful modeling capabilities of the pre-trained LLM, only a small portion of the parameters needs to be fine-tuned to improve CSI recovery accuracy with low training cost. Furthermore, a multiobjective loss function is designed to simultaneously optimize normalized mean square error (NMSE) and spectral efficiency (SE). Simulation results show that LLM4FB outperforms existing methods across various compression ratios and mobility levels, achieving high-precision CSI feedback with minimal computational capability from terminal devices. Therefore, LLM4FB presents a highly promising solution for next-generation wireless sensor networks and industrial IoT applications, where terminal devices are often strictly constrained by energy and hardware resources. Full article

Switching to electric vehicles (EVs) could help reduce air pollution in cities. This is especially important for cities in Russia that have grown quickly because of industry, like those in Siberia, where environmental problems are particularly acute. However, several factors continue to hinder the rapid expansion of EVs on the market, such as an additional strain on the energy infrastructure, which threatens to cause power outages. This study proposes a model for estimating the electricity consumption by EVs in the largest Russian cities, taking into account the technical characteristics of the EV fleet and climatic conditions. The calculations indicate that if 15% of the current car fleet are replaced by EVs, electricity consumption in the 16 largest cities in Russia would increase by 2.2 TWh per year in total. The estimated additional demand in particular cities varies between 33 mln and 769 mln kWh per year, depending on the number of vehicles and the local climate. Furthermore, we conducted an intra-day simulation of electricity consumption from EVs in a conditional Russian city with a population of over one million people. Three scenarios for the power grid load have been developed: (A) the maximum scenario, in which all EVs have a battery level of 0%; (B) the medium scenario, where EVs’ state of charge is distributed between 0% and 100%, and (C) the minimum scenario, involving charging scheduling that allows only EVs with a battery level of 20% or less to charge. The findings show that replacing just 15% of the car fleet with electric vehicles will trigger an increase in current daily household urban consumption of 28.4% in scenario (C), 75.6% in scenario (B) and 141.8% in scenario (A). Consequently, even in Russia’s largest cities, the further proliferation of EVs requires large-scale investments in power infrastructure. An additional 1 mln kWh used by EVs per day may require $160.7 mln investments in energy facilities and urban distribution networks. These findings highlight the necessity of a more thorough cost–benefit analysis of widespread electric vehicle adoption in densely populated urban areas. Full article

Printed circuit boards (PCBs) are critical in the electronics industry. As PCB layouts grow increasingly complex, defect detection processes often encounter challenges such as low image contrast, uneven brightness, minute defect sizes, and irregular shapes, making it difficult to achieve rapid and accurate automated inspection. To address these challenges, this paper proposes a novel object detector, YOLO-UMS, designed to enhance the accuracy and speed of PCB surface defect detection. First, a lightweight plug-and-play Unified Multi-Scale Feature Fusion Pyramid Network (UMSFPN) is proposed to process and fuse multi-scale information across different resolution layers. The UMSFPN uses a Cross-Stage Partial Multi-Scale Module (CSPMS) and an optimized fusion strategy. This approach balances the integration of fine-grained edge information from shallow layers and coarse-grained semantic details from deep layers. Second, the paper introduces a lightweight RG-ELAN module, based on the ELAN network, to enhance feature extraction for small targets in complex scenes. The RG-ELAN module uses low-cost operations to generate redundant feature maps and reduce computational complexity. Finally, the Adaptive Interaction Feature Integration (AIFI) module enriches high-level features by eliminating redundant interactions among shallow-layer features. The channel-priority convolutional attention module (CPCA), deployed in the detection head, strengthens the expressive power of small target features. The experimental results show that the new UMSFPN neck can help improve the AP50 by 3.1% and AP by 2% on the self-collected dataset PCB-M, which is better than the original PAFPN neck. Meanwhile, UMSFPN achieves excellent results across different detectors and datasets, verifying its broad applicability. Without pre-training weights, YOLO-UMS achieves an 84% AP50 on the PCB-M dataset, which is a 6.4% improvement over the baseline YOLO11. Comparing results with existing target detection algorithms shows that the algorithm exhibits good performance in terms of detection accuracy. It provides a feasible solution for efficient and accurate detection of PCB surface defects in the industry. Full article

The global shipping industry is surging ahead, and with it, a quiet revolution is taking place on the water: marine lithium-ion batteries have emerged as a crucial clean energy carrier, powering everything from ferries to container ships. When these vessels dock, they increasingly rely on shore power charging systems to refuel—essentially, plugging in instead of idling on diesel. But predicting how much power they will need is not straightforward. Think about it: different ships, varying battery sizes, mixed charging technologies, and unpredictable port stays all come into play, creating a load profile that is random, uneven, and often concentrated—a real headache for grid planners. So how do you forecast something so inherently variable? This study turned to the Monte Carlo method, a probabilistic technique that thrives on uncertainty. Instead of seeking a single fixed answer, the model embraces randomness, feeding in real-world data on supply modes, vessel types, battery capacity, and operational hours. Through repeated random sampling and load simulation, it builds up a realistic picture of potential charging demand. We ran the numbers for a simulated fleet of 400 vessels, and the results speak for themselves: load factors landed at 0.35 for conventional AC shore power, 0.39 for high-voltage DC, 0.33 for renewable-based systems, 0.64 for smart microgrids, and 0.76 when energy storage joined the mix. Notice how storage and microgrids really smooth things out? What does this mean in practice? Well, it turns out that Monte Carlo is not just academically elegant, it is practically useful. By quantifying uncertainty and delivering load factors within confidence intervals, the method offers port operators something precious: a data-backed foundation for decision-making. Whether it is sizing infrastructure, designing tariff incentives, or weighing the grid impact of different shore power setups, this approach adds clarity. In the bigger picture, that kind of insight matters. As ports worldwide strive to support cleaner shipping and align with climate goals—China’s “dual carbon” ambition being a case in point—achieving a reliable handle on charging demand is not just technical; it is strategic. Here, probabilistic modeling shifts from a simulation exercise to a tangible tool for greener, more resilient port energy management. Full article

Achieving high reliability remains the critical challenge for pulsed power supplies (PPS), whose core components are susceptible to severe degradation and catastrophic failure due to long-term operation under electrical, thermal and magnetic stresses, particularly those associated with high voltage and high current. This reliability challenge fundamentally limits the widespread deployment of PPSs in defense and industrial applications. This article provides a comprehensive and systematic review of the reliability challenges and recent technological progress concerning PPSs, focusing on three hierarchical levels: component, system integration, and extreme operating environments. The review investigates the underlying failure mechanisms, degradation characteristics, and structural optimization of key components, such as energy storage capacitors and power switches. Furthermore, it elaborates on advanced system-level techniques, including novel thermal management topologies, jitter control methods for multi-module synchronization, and electromagnetic interference (EMI) source suppression and coupling path optimization. The primary conclusion is that achieving long-term, high-frequency operation depends on multi-physics field modeling and robust, integrated design approaches at all three levels. In summary, this review outlines important research directions for future advancements and offers technical guidance to help speed up the development of next-generation PPS systems characterized by high power density, frequent repetition, and outstanding reliability. Full article

The detection of trace nitrogen dioxide (NO₂) is critical for environmental monitoring and industrial safety. Among various sensing technologies, chemiresistive sensors based on semiconducting metal oxides are prominent due to their high sensitivity and fast response. However, their application is hindered by inherent limitations, including low selectivity and elevated operating temperatures, which increase power consumption. Two-dimensional metal oxysulfides have recently attracted attention as room-temperature sensing materials due to their unique electronic properties and fully reversible sensing performance. Meanwhile, their combination with optoelectronic gas sensing has emerged as a promising solution, combining higher efficiency with minimal energy requirements. In this work, we introduce non-layered 2D indium oxysulfide (In₂S_xO_3−x) synthesized via a two-step process: liquid metal printing of indium followed by thermal annealing of the resulting In₂O₃ in a H₂S atmosphere at 300 °C. The synthesized material is characterized by a micrometer-scale lateral dimension with 6.3 nm thickness and remaining n-type semiconducting behavior with a bandgap of 2.53 eV. It demonstrates a significant response factor of 1.2 toward 10 ppm NO₂ under blue light illumination at room temperature. The sensor exhibits a linear response across a low concentration range of 0.1 to 10 ppm, alongside greatly improved reversibility, selectivity, and sensitivity. This study successfully optimizes the application of 2D metal oxysulfide and presents its potential for the development of energy-efficient NO₂ sensing systems. Full article