Search Results (7,877)

During urbanization, human activities have induced significant transformations in land use, leading to a huge imbalance in economic growth and nature development, posing severe threats to ecosystems. How to construct a stable and continuous ecological security pattern (ESP) in vulnerable areas like the Liaohe River Basin in Liaoning section has become a crucial challenge for regional management while facing the constraints of habitat fragmentation and the loss of landscape connectivity on sustainable development. Most research on ESP mainly relies on current situation or single scenario predictions, this study developed a “current assessment-future prediction-pattern optimization” framework. By simulating and comparing four distinct policy-oriented scenarios for 2030—Natural Development (ND), Cropland Protection (CP), Ecological Protection (EP), and Ecosystem-Service-Importance-Based Sustainable Development (ESIS)—this study aims to: (1) reveal the differentiated impacts of various policy orientations on future land use patterns; (2) compare the spatial evolution of ecological sources, resistance surfaces, ecological corridors, and key nodes between 2020 and under each 2030 scenario; and (3) synthesize an optimized ESP. This ESP is designed to balance economic and ecological needs, ultimately providing a scientific basis for watershed management. Furthermore, based on the simulation results, we propose a spatially explicit and adaptive management strategy termed the “one ribbon, two zones” pattern to guide the implementation of the optimized ESP within the basin. Full article

The assessment of reliability in non-repairable subsystems of mining electronic equipment represents a computationally challenging problem, particularly for complex and highly connected structures. This study presents a systematic comparative analysis of several deterministic approaches for reliability estimation, focusing on their computational efficiency, accuracy, and applicability. The investigated methods include classical boundary techniques (minimal paths and cuts), analytical decomposition based on the Bayes theorem, the logic–probabilistic method (LPM) employing triangle–star transformations, and the algorithmic Structure Convolution Method (SCM), which is based on matrix reduction of the system’s connectivity graph. The reliability problem is formally represented using graph theory, where each element is modeled as a binary variable with independent failures, which is a standard and practically justified assumption for power electronic subsystems operating without common-cause coupling. Numerical experiments were carried out on canonical benchmark topologies—bridge, tree, grid, and random connected graphs—representing different levels of structural complexity. The results demonstrate that the SCM achieves exact reliability values with up to six orders of magnitude acceleration compared to the LPM for systems containing more than 20 elements, while maintaining polynomial computational complexity. Qualitatively, the compared approaches differ in the nature of the output and practical applicability: boundary methods provide fast interval estimates suitable for preliminary screening, whereas decomposition may exhibit a systematic bias for highly connected (non-series–parallel) topologies. In contrast, the SCM consistently preserves exactness while remaining computationally tractable for medium and large sparse-to-moderately dense graphs, making it preferable for repeated recalculations in design and optimization workflows. The methods were implemented in Python 3.7 using NumPy and NetworkX, ensuring transparency and reproducibility. The findings confirm that the SCM is an efficient, scalable, and mathematically rigorous tool for reliability assessment and structural optimization of large-scale non-repairable systems. The presented methodology provides practical guidelines for selecting appropriate reliability evaluation techniques based on system complexity and computational resource constraints. Full article

Natural gas separation and purification are critical stages for ensuring product quality, operational safety, and economic efficiency in the energy sector. However, a significant research gap exists: conventional control systems, predominantly based on a proportional-integral-derivative (PID) controller, are often static and lack the adaptability required to handle fluctuations in raw gas composition and operating conditions. This review aims to systematically analyze modern control strategies to identify the most influential parameters and effective methodologies for enhancing process efficiency. The methods involve a comparative assessment of classical PID control against advanced intelligent approaches, including adaptive control, fuzzy logic, and machine learning (ML) models, based on a synthesis of the recent literature and industrial case studies. The key finding is that data-driven and intelligent methods (e.g., neural networks, adaptive fuzzy controllers) demonstrate superior performance in achieving precise parameter adjustment, improving responsiveness, and optimizing energy consumption compared to traditional static systems. Such an integrated strategy transforms decision-making into a multivariable optimization framework with objectives encompassing minimizing pollutants, lowering energy usage, and enhancing end-product specifications. The present work argues for employing methodologies like systemic analyses and advanced computational techniques—particularly artificial neural networks—to forecast gas stream attributes. Full article

Growing wastewater volumes and intensifying water scarcity are driving the need for affordable, sustainable solutions that enable safe urban water reuse and strengthen climate resilience. Policy frameworks such as SDG6 and EU water reuse requirements highlight that reclaimed water must meet strict environmental and public health standards. In contrast, conventional biological treatment cannot fully remove many emerging contaminants, underscoring the need for advanced treatment approaches that consistently deliver high-quality reclaimed water. In this context, this review examines the role of membrane technologies (MF, UF, NF, RO, FO) and membrane bioreactors (MBRs) in providing safe water in urban environments and in enhancing urban climate resilience, including decentralized systems and advanced reclamation needs. It also discusses the contribution of membrane-based solutions to sustainable cooling systems and heat-stress mitigation, as well as the integration of membrane technologies into green infrastructure and nature-based solutions for climate adaptation. Technical and economic performance is shaped by fouling, cleaning requirements, and energy use, making life-cycle and operational optimization critical for long-term sustainability. Case studies and EU-funded initiatives demonstrate that, with appropriate governance and design, membrane-based approaches can enable reliable reclaimed water supply, enhance water security, and contribute to circular urban water management. The analysis was based on peer-reviewed open-access publications, which may introduce a degree of selection bias. Full article

Large Language Models (LLMs) have emerged as a dominant paradigm in natural language processing, demonstrating strong performance across a wide range of generation and reasoning tasks. These systems depend on multi-stage training pipelines that integrate large-scale self-supervised pre-training, supervised fine-tuning, and alignment techniques. This paper presents a systematic mapping study of contemporary LLM training methodologies, emphasizing transformer-based architectures, optimization objectives, and data curation strategies as well as emerging sparse architectures such as Mixture-of-Experts (MoE) models. We analyze parameter-efficient fine-tuning approaches, retrieval-augmented generation frameworks, and multimodal training techniques, which we organize into a unified comparative taxonomy. We discuss key technical challenges such as scalability constraints, hallucination, bias amplification, and alignment–capability tradeoffs, then identify emerging research directions such as reasoning-centric training. This work provides a concise technical reference for researchers and practitioners working on scalable and reliable language model training. Full article

Large language models (LLMs) with the Transformer architecture as the core have made significant progress in the field of natural language processing, and their application value in English teaching has also attracted much attention. In tasks such as text generation, question-answering systems, and translation, the processing capabilities of LLMs have significantly improved. However, existing LLMs have problems such as insufficient coverage of professional knowledge, rough semantic parsing, and weak personalized services. To address the aforementioned issues, this study proposes a dual-path retrieval-enhanced generation scheme that integrates vector databases and intelligent agents, aiming to improve the application of large models in English language teaching. Semantic retrieval of unstructured data in English teaching is realized through vector databases, knowledge is dynamically acquired by combining agents, and the accuracy is improved by using Bloom filters to fuse dual-path retrieval. At the same time, the retrieval efficiency is optimized by an importance-oriented algorithm, and user profiles are constructed based on multi-dimensional data to achieve personalized adaptation. Experiments show that the maximum optimization of the retrieval time of this scheme can reach 26.32%, and the highest retrieval accuracy can reach 86%. The key indicators and scores in tasks such as English knowledge retrieval and question-answering reasoning are better than those of the comparative schemes, providing an effective technical path for intelligent English teaching. Full article

Urban blue spaces, including rivers, lakes, and ponds, are increasingly recognized as nature-based solutions for mitigating the Urban Heat Island (UHI) effect. However, fragmented evidence and inconsistent evaluation frameworks have limited their effective integration into climate-adaptive urban planning. This study conducts a comprehensive bibliometric analysis and systematic review to synthesize current knowledge on the cooling effects of urban blue spaces. A total of 110 peer-reviewed publications published between 2015 and 2025 were retrieved from the Web of Science Core Collection and analyzed using the Bibliometric-Systematic Literature Review (B-SLR) framework. The results reveal a rapidly growing research field characterized by increasing interdisciplinary integration. Evidence consistently indicates that the cooling effects of blue spaces exhibit pronounced diurnal and seasonal variability, highlighting a “diurnal paradox” of daytime cooling versus nighttime warming risks, with stronger impacts in summer than in winter. Cooling performance is governed by non-linear morphological thresholds regarding size, shape, spatial configuration, and upwind location, where aerodynamic ventilation is critical for extending the cooling range. Moreover, the interaction between blue spaces, building morphology (gray infrastructure), and green infrastructure plays a decisive role: specific density thresholds in built environments can constrain cooling diffusion, whereas synergistic blue–green integration significantly enhances thermal regulation through coupled evaporative, shading, and ventilation processes. Overall, this review demonstrates a clear shift from isolated temperature-based assessments toward systemic, planning-oriented approaches emphasizing multi-scale integration and context-sensitive design. The findings provide operational parameters and demand-based strategies for optimizing blue infrastructure in climate-resilient urban planning. Full article

The hardware power-aware keyword spotting (KWS) implementation requires small memory footprint, low-complex computation, and high accuracy performances. Binary neural networks (BNNs) naturally satisfy these constraints. They quantize both weights and activations to 1-bit. This reduces storage and replaces most multiply–accumulate operations with bitwise operations. However, such extreme quantization incurs substantial information loss and leaves a noticeable accuracy gap relative to full-precision models. Optimization is also more difficult because the sign function is non-differentiable, and surrogate-gradient updates introduce gradient mismatch. To preserve the hardware benefits of BNNs while alleviating the accuracy degeneration induced by 1-bit quantization, this article addresses the problem from two complementary aspects: Firstly, a Stepwise Self-Knowledge Distillation (SSKD) training approach is proposed to effectively improve the student BNN’s accuracy performance. The SSKD training framework achieves effective supervision for student BNNs. A Stepwise Training Strategy is proposed to optimize the training stability and accuracy. Weight Scaling Factor improves the student’s representational capability. Secondly, an extremely lightweight Binary Temporal Convolutional ResNet (BTC-ResNet) is also proposed. The parameters and calculations inside the network are greatly reduced for the inference. Experiments on the GSCD v1 and GSCD v2 benchmarks demonstrate the effectiveness of our methods for low-power keyword spotting. For the 12-class task, BTC-ResNet14 achieves 97.23% accuracy on GSCD v1 and 97.31% on GSCD v2 with 0.75 Mb parameters and 1.35 M FLOPs. For the 35-class task on GSCD v2, it reaches 95.56% accuracy with 0.76 Mb parameters and 1.35 M FLOPs. These results indicate that our method achieves a competitive accuracy–efficiency balance relative to recent distillation-based BNN KWS baselines reported in the comparative experiments. All these studies are helpful and promising for future KWS deployment on low-power hardware devices. Full article

Flat spots on railway wheels critically threaten operational safety, accelerating track damage, noise pollution, and energy waste. Their repetitive, high-magnitude impacts dissipate mechanical energy as ground vibration and noise, directly reducing traction efficiency. This paper presents a comprehensive review of recent vibration and acoustic detection methods, comparing onboard and stationary wayside systems. The literature from 2019 to 2025 shows a trend toward machine learning and deep learning, often reaching nearly 100% accuracy in laboratory or simulated settings. However, most studies lack real-world validation with naturally occurring defects. Bridging this gap requires industry–academic collaboration for operational data and testing. Crucially, systems must classify defect severity in line with maintenance thresholds rather than focus on minor, non-actionable faults. Integrating these technologies into condition-based maintenance and predictive digital twins will enable optimization of scheduling and work orders. Future efforts should leverage edge computing for real-time analysis, federated learning for data scarcity, and energy harvesting for sensor autonomy. The goal is to develop field-validated, integrated systems that enhance safety, reduce energy loss, and improve reliability. Full article

Background: Natural disasters can trigger hazardous material (Hazmat) releases and damage valuable assets, increasing human, environmental, and economic losses. Effective pre-disaster relocation planning is therefore critical but operationally challenging. Methods: This study develops a mixed-integer programming model for the pre-disaster relocation of Hazmat and valuable assets (HVAs). The model jointly optimizes safe-location activation, fleet allocation, and trip scheduling within a limited warning period, subject to vehicle availability, storage and capacity limits. The objective minimizes total cost, including facility activation, transportation, and penalties for unrelocated inventories. The model is solved using the Gurobi Optimizer. A base scenario and sensitivity analyses on fleet size and safe-location capacity are conducted using data from a cyclone-prone region in Oman. Results: In the base scenario, 73.4% of HVAs are relocated by activating 10 safe locations. Sensitivity analysis shows rapid gains at small fleet sizes, followed by diminishing returns beyond a threshold. Over 95% of HVAs are relocated by doubling safe-location capacities with 80 vehicles or tripling capacities with 65 vehicles. Conclusions: Total vehicle capacity, time-window, and safe-location capacity constraints become binding at different thresholds, highlighting the need for balanced investments. The proposed model provides an analytics-driven decision-support tool for risk-aware, time-bounded disaster relocation planning. Full article

Autonomous flight is a critical capability for unmanned aerial vehicles (UAVs), enabling applications in wildlife and plant protection, infrastructure inspection, search and rescue, and other complex missions. Although some learning-based methods have achieved considerable progress, traditional algorithms still struggle with real-world challenges, due to the partially observable nature of environments and limited experience regarding the properties of dynamic unknown environments where threats and targets are movable and unpredictable. To address these difficulties, it is necessary to achieve autonomous guidance for UAVs performing long-range missions in dynamic environments (LRGDEs), and to develop a novel end-to-end algorithm that can overcome partial observability under limited state transitions. In this paper, we propose an RNN-enhanced Diverse Curriculum-driven Learning Algorithm (REDCRL) based on deep reinforcement learning. We modify the structure of traditional actor–critic networks and introduce Bi-LSTM into policy networks (referred to as Bi-LSTM-modified Policy Networks (BLPNs)) to alleviate observation incompleteness. Furthermore, to fully exploit the potential value of data and mitigate the problem of insufficient samples, we develop an Adaptive Multi-Feature Evaluation Experience Replay (AMFER) method to reshape the process of experience replay buffer construction and sampling. In addition, the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm is adopted to optimize UAV-maneuver decision policies. Compared with traditional algorithms, the proposed algorithm can accelerate policy convergence and improve the performance of the trained policy. Full article

This research investigates the dynamic response of three-dimensional concrete beams with coated aggregates subjected to transient wave propagation induced by piezoelectric actuators (PZT-lead zirconate titanate). The aim was to assess the effects of coated aggregates on concrete’s damping behavior and wave attenuation. The effect of various coating materials (epoxy and rubber) and replacement levels ranging from 5% to 25% by volume of the natural coarse aggregates on wave attenuation and energy dissipation were investigated through Finite Element Modeling (FEM) using Abaqus/CAE 6.14-1 software. Moreover, the effect of coating thickness is investigated for thicknesses ranging from 1.0 mm to 3.5 mm. The findings show that the replacement level, coating thickness, and coating material have a major effect on damping characteristics of concrete. It was also observed that rubber-coated aggregates with a 3.0 mm coating thickness (20% replacement level) exhibited an optimal damping ratio of 6.15%, representing a 29.5% increase, and offer enhanced energy dissipation, better damping performance, and the ability to alter wave travel paths, all of which could be advantageous for Structural Health Monitoring (SHM) applications. In line with this, the damping ratio of concrete beam models with epoxy- and rubber-coated aggregates, embedded with PZT material, was significantly higher (by approximately 1% to 18%) compared to that of the concrete model without PZT materials. Additionally, the findings showed that the concrete’s damping properties were greatly impacted by the interaction between PZT materials and coated aggregates. All things considered, the dynamic response and damping performance of concrete with coated-aggregate surface properties were successfully assessed using PZT-based simulations. Full article

This article presents a systematic synthesis of variationally grounded approaches for the design and optimization of mining structural infrastructure. This study is motivated by the critical need to ensure stiffness, reliability, and operational availability under severe loading, mass constraints, and aggressive environmental conditions. Methodologically, the study situates structural modeling and synthesis within the continuity of the principle of stationary action. It demonstrates that, in the quasi-static regime, structural equilibrium is obtained as the stationarity of the total potential energy; consequently, the finite element method (FEM) arises naturally as a Ritz–Galerkin approximation of this underlying variational statement. On this basis, topology optimization is interpreted as a physics-constrained optimization problem wherein the design is posed as an outer optimality level acting over an energetically defined state. It is worth noting that SIMP-based formulations require explicit regularization to define the effective problem being solved. Emphasis is placed on the traceability between physical assumptions, discretization choices, regularization, and the resulting structural interpretations. The core contribution of this paper is a systematic literature review that consolidates evidence across variational mechanics, FEM-based optimization, and industrial applications, identifying recurrent methodological patterns and gaps that currently limit transfer to mining practice. Furthermore, a fully specified illustrative case is included to demonstrate reporting discipline and methodological consistency, rather than as a validation of a new optimization method. The conclusions highlight that a variational reading provides a coherent theoretical backbone for structural analysis, synthesis, simulation, and physics-based digital twins, while also clarifying the extensions required for industrial deployment, such as stability constraints, manufacturability, and multiphysics coupling within Mining 4.0 workflows. Full article

The rapid development of medical technologies requires architects to implement a future-proofing approach while designing medical facilities, despite the inherent uncertainty of long-term change. This challenge is particularly visible within hospital emergency departments (HEDs), which play a critical role as first-contact units and life-saving infrastructures. Due to their specific function, HEDs are a challenging environment for implementing new solutions, as they rely on proven frameworks designed to ensure continuity of care and operational efficiency. This raises the key question: how can modern technologies and architectural strategies streamline workflows in HEDs without overwhelming medical staff? Considering current challenges, an equally important factor in the development of emergency departments is their preparedness for crisis situations, such as pandemics, war threats and natural disasters. How can architectural design enable the implementation of given design strategies, aiming to ensure opportunities for development while simultaneously preparing for all-hazard scenarios? The authors gathered existing trends and solutions aimed at preparing hospital emergency departments for future challenges: positive/neutral, such as technological development, but also negative, such as currently ongoing war threats or risk of the next pandemic. Despite the apparent thematic extremity, certain systematic architectural solutions using a transdisciplinary approach may be the answer to these occurrences. The mentioned architectural solutions and factors were synthesized and subjected to design-oriented review based on existing case studies of a few Polish hospitals, which are simultaneously studied as case studies for broader doctoral research in the field of effectiveness assessment. The selected Polish hospital emergency departments are used as an illustrative, analytical reference to support the interpretation and synthesis of the reviewed literature. The contextual analysis enables the identification of transferable, design-oriented strategies relevant to broader emergence medicine architecture and applicable within European units. Examples from Polish units in particular are used as reference and background for discussion, rather than as empirical case studies. The study provides an overview of contemporary and future-oriented solutions in hospital architecture, focusing on the impact and feasibility within the hospital emergency departments. The synthesis highlights the importance of designing flexible spaces prepared for future technological advances, such as oversized service shafts, increased floor heights, and modular layouts. Additionally, the study focuses on the spatial connotations of emerging technologies like medical robotics, their maintenance areas and possible challenges. All of this is interrelated to social, demographic, and economic trends. These include the development of hospital networks, the evolving patient profile, inter-hospital information flow, and the growing role of highly specialized medical units. In terms of rapid challenges like wars or armed threats, factors revealed within the review indicate levels of HED readiness to face the conflict, mainly in terms of surge capacity but also structural durability and reserve resources. The post-pandemic context, in turn, assumes rapid expansion of the hospital into temporary and flexible structures and reversible zoning allowing for patient segregation and separation. Together, these insights outline pathways for creating resilient, adaptable, and efficient emergency care environments resilient to unforeseen challenges. Considering future scenarios of emergency departments, two main scenarios were identified: “the hospital of the future”, continuing overall development and adapting to rapid technological innovations, and “the crisis-resilient hospital”, resistant to various crisis scenarios, such as pandemics or war threats. The optimal development of the unit assumes both openness to technological changes and preparation of key zones for all-hazard scenarios. This review aims to synthesize architectural implications of technological and socio-demographic changes, not to provide a full empirical study. Adopting an exploratory framework, the review refers to technological innovations and crisis preparedness as external drivers shaping the spatial organization of hospital emergency departments and their adaptability to future challenges. Because of various inhibitors (economic, political, hierarchical), not all hospitals can introduce the described improvements, but the synthesis may serve as a knowledge source for future investments. The review was also conducted to support design decisions under conditions of uncertainty. The choice to address all the external factors collectively was induced to provide transferability of solutions and coherence of possible scenarios, which may happen simultaneously. Full article

Mining-resource-based cities, as distinctive human–environment systems, face urgent challenges from intensified urbanization and mining, leading to land imbalance and ecosystem service degradation. To enhance resilience, it is essential to identify the evolution and drivers of ecosystem services and construct targeted ecological compensation models. This study focuses on Xinzhou, a representative mining city in China, and systematically analyzes three aspects: (1) spatiotemporal dynamics of land use and ecosystem service value (ESV) from 2000 to 2023 using Markov chains, equivalent factor method, hotspot and sensitivity analyses; (2) identification of ESV driving mechanisms through an integrated “stepwise regression + geographical detector” framework; and (3) formulation of ecological compensation models via quantification of priority indices, demand intensity coefficients, and compensation standards. Key findings indicate that land conversion was concentrated in coalfield zones and surrounding built-up areas, involving 2,518,341.75 hm² (35.76% of total area), primarily characterized by a reduction in farmland and expansion of forest, grassland, and construction land. ESV showed a striped spatial pattern, with higher values in mountainous zones and lower values in valleys and basins with frequent human activity. The northwest coalfield region experienced an initial decline followed by a recovery in ESV. Annual mean temperature emerged as the dominant driver, while DEM influence increased annually. All factor interactions exhibited synergistic effects, with natural variables exerting greater influence than socio-economic ones. Ecological compensation demand was high overall, especially in Wutai, Kelan, and Pianguan counties, with high-value compensation areas mainly distributed in the eastern and central parts of Xinzhou. Looking ahead, a compensation framework prioritizing ecological–economic optimization should be developed, guided by zoned, typological, and dynamic configurations. By analyzing ecosystem governance from the perspective of a mining-resource-based city, this study enhances global ecosystem service evaluation frameworks and offers a replicable model to advance transnational ecological cooperation and green urban transformation. Full article