Search Results (884)

Machine learning (ML) has become a cornerstone of critical applications, but its vulnerability to data poisoning attacks threatens system reliability and trustworthiness. Prior studies have begun to investigate the impact of data poisoning and proposed various defense or evaluation methods; however, most efforts remain limited to quantifying performance degradation, with little systematic comparison of internal behaviors across model architectures under attack and insufficient attention to interpretability for revealing model vulnerabilities. To tackle this issue, we build a reproducible evaluation pipeline and emphasize the importance of integrating robustness with interpretability in the design of secure and trustworthy ML systems. To be specific, we propose a unified poisoning evaluation framework that systematically compares traditional ML models, deep neural networks, and large language models under three representative attack strategies including label flipping, random corruption, and adversarial insertion, at escalating severity levels of 30%, 50%, and 75%, and integrate LIME-based explanations to trace the evolution of model reasoning. Experimental results demonstrate that traditional models collapse rapidly under label noise, whereas Bayesian LSTM hybrids and large language models maintain stronger resilience. Further interpretability analysis uncovers attribution failure patterns, such as over-reliance on neutral tokens or misinterpretation of adversarial cues, providing insights beyond accuracy metrics. Full article

This work presents a highly compact and scalable 1.2-kV SiC MOSFET half-bridge building-block module enabled by a die-integrated 3D PCB packaging technology. Compared with conventional DBC-based or TO-247-based SiC half-bridge modules, the proposed design reduces the physical volume and weight by more than 90% while maintaining full compatibility with standard PCB manufacturing processes. The vertically laminated DC+/DC− conductors and symmetric PCB–die–PCB stack establish a tightly confined commutation loop, resulting in a measured power-loop inductance of 2.2 nH and a 3.8 nH gate-loop inductance—representing up to 94% and 89% reduction relative to discrete device implementations. Because the parasitic parameters are intrinsically well-balanced across replicated units and the mutual inductance between adjacent modules remains extremely small, the structure naturally supports current sharing during parallel operation. Thermal and insulation evaluations further confirm the suitability of copper filling via high-Tg laminated PCB substrates for high-power SiC applications, achieving withstand voltages exceeding twice the rated bus voltage. The proposed module is experimentally validated through finite-element parasitic extraction and 950 V double-pulse testing, demonstrating controlled dv/dt behavior and robust switching performance. This work establishes a manufacturable and parallel-friendly packaging approach for high-density SiC power conversion systems. Full article

With the large-scale integration of new energy and power electronic devices into power systems, frequency stability has become an increasingly critical concern. To maintain frequency stability while mitigating the high capital expenditure of energy storage systems (ESSs), this paper develops a control framework centered on edge energy management terminals (EEMTs). The design is based on a demonstration project in which distributed energy resources (DERs) and flexible loads collaboratively provide frequency regulation. A monitoring station is implemented to make fast frequency response (FFR) resources dispatchable, detectable, measurable, and tradable. Furthermore, a control strategy tailored for building- and factory-level applications is proposed. This strategy enables real-time optimal scheduling of DERs and flexible loads through coordinated communication between EEMTs and net load units (NLUs). Two field tests further demonstrate the effectiveness and advantages of the proposed approach. In addition, this paper proposes a coordinated scheme in which wind farms and NLUs jointly participate in frequency regulation, aiming to mitigate the response delay of NLUs and the secondary frequency drop observed in wind farms. The feasibility and benefits of this scheme are validated through experimental tests. Full article

An effective housing policy must ensure affordability for individuals across all income levels by integrating advanced technological innovations with comprehensive socioeconomic strategies. Affordable housing fosters social inclusion, whereas sustainability supports long-term environmental protection and economic stability. The success and long-term sustainability of affordable housing initiatives are heavily influenced by current socioeconomic conditions, emphasizing the need for context-specific, inclusive, and sustainable housing solutions. Benchmarks are crucial in affordable housing to determine if it is climate-positive, aligning with the goals of the United Nations’ Sustainable Development Goal 11.1, which seeks to provide affordable and sustainable housing for everyone by 2030. This study uses the Scopus database to perform a scientometric analysis of 595 publications (2015–2024) on sustainability and affordability in housing. Using R-Studio 2025.05.1 + 513.pro3 and VOSviewer 1.6.20, it examines bibliographic trends, research gaps, and collaboration patterns across countries and journals. This study highlights performance thresholds related to economic, environmental, energy, territorial, and climatic factors. However, cost and ecological objectives can cause conflict with each other practically, and hence a balanced approach including green practices, efficient materials, and subsidies is crucial. There is a need for policymakers to address market gaps to prevent socially exclusive or environmentally harmful outcomes, maintain long-term urban resilience, and ensure sustained urban resilience and equitable access to affordable, sustainable housing by 2030. Integrating sustainable materials, circular and climate-resilient design, smart technologies, inclusive governance, and evidence-based policies is crucial for advancing affordable, equitable, and resilient housing. This approach guides future research and policy toward long-term social, economic, and environmental benefits. The findings and recommendations promote sustainable, affordable housing, emphasizing the need for further research on climate-resilient, energy-efficient, and cost-effective building solutions. Full article

The building sector represents a major source of global carbon emissions, with heating and cooling systems being particularly critical contributors, making the evaluation of sustainable low-carbon alternatives an urgent priority. In this study, life cycle assessment (LCA) methodology is used to analyze ground source heat pump (GSHP) systems against traditional heating, ventilation, and air conditioning (HVAC) systems based on project data from the city of Jinan and electrical grid characteristics of Northern China. It is specified that the functional unit is providing heating and cooling that maintains the indoor temperature of the building between 18 °C and 26 °C for 20 years. Following ISO 14040 standards, carbon emissions and economic performance across four phases—production, transportation, construction, and operation—over a 20-year life cycle were quantified using actual material inventory data and region-specific carbon emissions factors. The results demonstrate obvious environmental advantages for GSHP systems, which achieve a 51% reduction in life cycle carbon emissions compared to traditional systems based on the current power generation structure. Furthermore, sensitivity analysis shows that as the proportion of renewable energy in the grid increases to meet carbon neutrality targets, the reduction potential can even reach 88%. Economic analysis reveals that despite higher initial investments, GSHP systems achieve favorable performance with a positive 20-year net present value and an acceptable dynamic payback period for the project. This study shows that GSHP systems represent a viable strategy for sustainable building design in northern China, and the substantial carbon reduction potential can be further enhanced through grid decarbonization and renewable energy integration. The implementation of the GSHP system in newly constructed buildings, which require both heating and cooling, in Northern China, can be an effective strategy for advancing carbon neutrality goals. Full article

Recognizing handwritten Arabic characters written by children via scanned or camera-captured images is a challenging task due to variations in writing style, stroke irregularity, and diacritical marks. Although deep learning has advanced this field, building reliable systems remains challenging. This study introduces a stacking ensemble framework for sensor-acquired handwriting data, enhanced with a dynamic confidence-thresholding mechanism designed to improve prediction reliability. The framework integrates three high-performing convolutional neural networks (ConvNeXtBase, DenseNet201, and VGG16) through a fully connected meta-learner. A key feature is the use of an optimized threshold that filters out uncertain predictions by maximizing the macro ${\mathrm{F}}_1$ score on validation data. The framework is evaluated on two benchmark datasets for children’s Arabic handwriting: Hijja and Dhad. The results demonstrate state-of-the-art performance, with an accuracy of $95.13\%$ and ${\mathrm{F}}_1$ score of $94.62\%$ on Hijja and an accuracy of $96.14\%$ and ${\mathrm{F}}_1$ score of $95.59\%$ on Dhad. Compared to existing methods, the proposed approach achieves more than a $3\%$ improvement in Hijja accuracy while maintaining robust performance across diverse character classes. These findings highlight the effectiveness of confidence-based stacking ensembles in enhancing reliability for Arabic handwriting recognition and suggest strong potential for automated educational assessment tools and intelligent tutoring systems. Full article

Building a sustainable and resilient humanitarian logistics system is essential for reducing disaster losses and supporting long-term socio-economic recovery. Following a major disaster, rapidly organizing casualty evacuation while maintaining system robustness is a fundamental component of sustainable emergency management. This study develops a two-stage robust optimization model for designing a sustainable humanitarian logistics network that simultaneously accounts for two critical post-disaster uncertainties: (i) interruption risks at temporary medical points and (ii) uncertain casualty demand. By explicitly differentiating deprivation costs between mild and serious injuries, the model quantifies human suffering in monetary terms, thereby integrating social and economic sustainability considerations into the optimization framework. A customized column-and-constraint generation (C&CG) algorithm with proven finite convergence is proposed to ensure tractability and practical applicability. Using the 2008 Wenchuan earthquake as a real-world case study, involving 10 affected areas and 10 candidate temporary medical points, the results demonstrate that the proposed approach yields evacuation plans that remain feasible under all tested worst-case realizations, substantially reducing deprivation costs compared with existing benchmarks. The findings highlight that strategically increasing the capacity of key temporary medical nodes enhances the sustainability and resilience of the emergency medical system, offering evidence-based insights for designing sustainable and robust disaster-response strategies. Full article

Recycling education is important for promoting pro-environmental sustainable behavior, yet traditional approaches often lack engagement and impact, particularly among younger audiences. This study presents a digital, turn-based card strategy game designed to teach recycling principles and concepts through interactive city-building mechanics. Set in an augmented reality environment, the game challenges players to balance population growth, resource use, and waste management to maintain a high well-being score for their city. Players construct digital buildings (houses, recycling facilities, resource infrastructures), each influencing waste production, recycling efficiency, and overall well-being. The game integrates educational content with engaging decision-making, aiming to foster system thinking and eco-conscious behavior. Unlike prior AR approaches, this game focuses on digital interaction, leveraging immersive game-based learning. Usability and engagement were evaluated using the in-game version of the Game Experience Questionnaire (GEQ). Findings support that users responded positively to the prototype’s game experience, suggesting that the digital game is promising. The study contributes to the growing field of digital pro-environmental education, providing insights into how interactive gameplay can support environmental awareness and laying groundwork for future evaluation of its educational impact. Full article

Building energy consumption is a major source of carbon emissions, with the heating energy demand of rural buildings in the hot summer and cold winter (HSCW) zone having increased 575-fold over the past 15 years. This research investigated an optimized solar–air source heat pump (SASHP) system to meet the heating demand of rural residences in this region. First, a typical rural building model was developed using SketchUp, and its heating load was simulated using TRNSYS, revealing an average load of 3.38 kW and a peak load of 5.9 kW. Based on the latest technical standards, the SASHP system was designed and simulated using TRNSYS, achieving an overall coefficient of performance (COP) of 3.67 while maintaining indoor thermal comfort within ISO 7730 Category II. Subsequently, the system was optimized through GenOpt to minimize the annual equivalent cost, yielding key parameters: a 15 m² solar collector at a 40.75° tilt, a 0.35 m³ water tank, and a 10.16 kW air source heat pump. Compared with the initial design, the optimized configuration achieved reductions of 35.60% in initial investment and 32.68% in annual equivalent costs. By ensuring thermal comfort and overcoming the economic barrier, this study provides a viable pathway for adoption and promotion of renewable heating technology in rural areas. Full article

With the increasing functional and geometric complexity of modern steel buildings, irregular-shaped beam-to-column joints are becoming common in engineering practice. However, their seismic behavior remains insufficiently understood, particularly for configurations with geometric asymmetry and complex stress transfer mechanisms. This study experimentally investigates the seismic performance of irregular steel-beam-to-concrete-filled steel tube (CFST) column joints incorporating inclined internal diaphragms (IIDs), taking unequal-depth beam (UDB) and staggered beam (SB) joints as representative cases. Two full-scale joint specimens were designed and tested under cyclic loading to evaluate their failure modes, load-bearing capacity, stiffness/strength degradation, energy dissipation capacity, strain distribution, and panel zone shear behavior. Both joints exhibited satisfactory strength and initial stiffness. Although diaphragm fracture occurred at approximately 3% drift, the joints retained 45–60% of their peak load capacity, based on the average strength of several loading cycles at the same drift level after diaphragm failure, and maintained stable hysteresis with average equivalent damping ratios above 0.20. Final failure was governed by successive diaphragm fracture followed by the tearing of the column wall, indicating that the adopted diaphragm thickness (equal to the beam flange thickness) was insufficient and that welding quality significantly affected joint performance. Refined finite element (FE) models were developed and validated against the test responses, reasonably capturing global strength, initial stiffness, and the stress concentration patterns prior to diaphragm fracture. The findings of this study provide a useful reference for the seismic design and further development of internal-diaphragm irregular steel-beam-to-CFST column joints. Full article

The intensification of urban heat islands in high-density coastal slope areas poses significant challenges to sustainable development. From the perspective of sustainable urban design, this study investigates adaptive greening strategies to mitigate thermal stress, aiming to elucidate the key microclimate mechanisms under the combined influence of sea breezes and complex terrain to develop sustainable solutions that synergistically improve the thermal environment and energy efficiency. Combining field measurements with ENVI-met numerical simulations, this research systematically evaluates the thermal impacts of various greening strategies, including current conditions, lawns, shrubs, and tree configurations with different canopy coverages and leaf area indexes. During summer afternoon heat episodes, the highest temperatures within the building-dense sites were recorded in unshaded open areas, reaching 31.6 °C with a UTCI of 43.95 °C. While green shading provided some cooling, the contribution of natural ventilation was more significant (shrubs and lawns reduced temperatures by 0.23 °C and 0.15 °C on average, respectively, whereas various tree planting schemes yielded minimal reductions of only 0.012–0.015 °C). Consequently, this study proposes a climate-adaptive sustainable design paradigm: in areas aligned with the prevailing sea breeze, lower tree coverage should be maintained to create ventilation corridors that maximize passive cooling through natural wind resources; conversely, in densely built areas with continuous urban interfaces, higher tree coverage is essential to enhance shading and reduce solar radiant heat loads. Full article

Multi-objective Bayesian Optimization (MOBO) has become a promising strategy for accelerating process development in Laser Powder Bed Fusion of Metals (PBF-LB/M), where experimental evaluations are costly, and design spaces are high-dimensional. This study investigates how different initialization strategies affect MOBO performance in a discrete, machine-limited parameter space during the fabrication of AZ31 magnesium alloy. Three approaches to constructing the initial experimental dataset—Latin Hypercube Sampling (V1), balanced-marginal selection (V2), and prior fractional-factorial sampling (V3)—were compared using two state-of-the-art MOBO algorithms, DGEMO and TSEMO, implemented within the AutoOED platform. A total of 180 samples were produced and evaluated with respect to two conflicting objectives: maximizing relative density and build rate. The evolution of the Pareto front and hypervolume metrics shows that the structure of the initial dataset strongly governs subsequent optimization efficiency. Variant V3 yielded the highest hypervolumes for both algorithms, whereas Variant V2 produced the most uniform Pareto approximation, highlighting a trade-off between global coverage and structural distribution. TSEMO demonstrated faster early convergence, whereas DGEMO maintained broader exploration of the design space. Analysis of duplicate experimental points revealed that discretization and batch selection can considerably limit the effective search diversity, contributing to an early saturation of hypervolume gains. The results indicate that, in constrained PBF-LB/M design spaces, MOBO primarily serves to validate and refine a well-designed initial dataset rather than to discover dramatically new optima. The presented workflow highlights how initialization, parameter discretization, and sampling diversity shape the practical efficiency of MOBO for additive manufacturing process optimization. Full article

Atrium spaces are widely applied in university buildings. However, achieving effective energy reduction while maintaining adequate daylighting and indoor comfort remains a major challenge at the early design stage. This study identifies key building form design variables significantly influencing atrium daylighting, energy use, and thermal comfort, including building orientation, atrium width-to-depth ratio, atrium aspect ratio, atrium bottom area ratio, and skylight–roof ratio. A multi-objective optimization (MOO) framework is proposed to balance daylight performance, energy consumption, and thermal comfort under fixed envelope parameters. Using typical single- and double-atrium teaching buildings in cold regions as case studies, this research adopts Useful Daylight Illuminance (UDI), Energy Use Intensity (EUI), and Discomfort Time Percentage (DTP) as key indicators to evaluate the interactions between design parameters and building performance. Based on the Pareto-optimal results for the studied prototypes, a south-by-west orientation, moderately slender atrium proportions, relatively compact atrium bottom areas, and medium skylight–roof ratios together yield a balanced performance. Compared with the reference to the initial solution, the optimized solutions reduce EUI by up to 5.66% while also improving UDI and DTP. These results are intended as quantitative references and optimization for early-stage geometric forms design of atrium teaching buildings in cold regions. Full article

With the widespread adoption of smartphones and wearable devices, localization systems have become increasingly important in modern society. While Global Positioning System (GPS) technology is widely accepted as a standard outdoors, accurately determining user location indoors remains a significant challenge despite extensive research efforts. Indoor positioning systems (IPSs) play a critical role in various sectors, including retail, tourism, transportation, healthcare, and emergency services. However, existing solutions require costly infrastructure deployments, complex area mapping, or offer suboptimal user experiences without achieving satisfactory accuracy. This paper introduces FloorTag, a scalable, low-cost, and minimally invasive hybrid IPS designed specifically for smartphone platforms. FloorTag leverages a combination of 2D visual markers placed on floor surfaces at key locations, and inertial sensor data from mobile devices. Each marker is associated with a unique identifier and precise spatial coordinates, enabling an immediate reset of accumulated localization errors upon detection. Between markers, a pedometer-based dead reckoning module maintains continuous location tracking. The localization process is designed to be seamless and unobtrusive to the user. When activated by the app during navigation, the phone’s rear camera, naturally angled toward the floor during walking, captures markers. This solution avoids explicit user scans while preserving the performance benefits of visual positioning. To model the indoor environment, FloorTag introduces the concept of Path-Points, which discretize the walkable space, and Informative Layers, which add semantic context to the navigation experience. This paper details the proposed methodology and the client–server system architecture and presents experimental results obtained from a prototype deployed in an academic building at the University of Catania, Italy. The findings demonstrate reliable localization at approximately 2 m spatial granularity and near-real-time performance across varying lighting conditions, confirming the feasibility of the approach and the effectiveness of the system. Full article

The main challenge for the scientific community is to mitigate climate change impacts while reducing energy consumption, without compromising comfort and quality of life. Buildings in hot climates require specific design strategies to limit the effects of extreme weather and heat waves. Standardized modern buildings, often unsuitable for hot and arid climates, lead to high energy consumption, mainly due to cooling systems, causing both discomfort and energy inefficiency. Previous studies have shown that solutions inspired by local vernacular architecture are often more effective than conventional construction techniques. This paper investigates the thermal response and discomfort intensity in two building models exposed to various climate scenarios: a typical modern residential building and a bioclimatic vernacular-inspired building. The analysis is conducted through dynamic thermal simulations under current as well as future medium- and long-term climate change scenarios. The study evaluates the buildings’ ability to adapt to future environmental changes, an aspect that has not yet been studied in depth. Results show that contemporary buildings experience significantly higher levels of thermal discomfort than vernacular buildings under both present (TMY) and future (SSP1-2.6 and SSP5-8.5, 2080) climate conditions. Results show that under the present climate, the vernacular building exhibits about 22% fewer discomfort hours than the contemporary one and roughly half the overheating integrated degree-hours. Under future scenarios, overheating increases by 25.8% to 67.7% in the contemporary building and 36.1% to 89.6% in the vernacular building, yet the vernacular building consistently maintains substantially lower discomfort levels. Overall, vernacular inspired envelopes remain more resilient to warming in all scenarios, but additional adaptation measures are required to ensure acceptable summer comfort by late century. Full article