Browse Articles
Discover research articles across all indexed journals
A knowledge graph and multi-agent reinforcement learning model for psychological risk identification and intervention in online learning
Abstract As the scale of online learning expanded, students tended to exhibit declining learning behaviors and accumulating task backlogs during self-directed study. These issues further led to reduced learning motivation and increased psychological pressure. To address this, this study constructed an innovative model that integrated a knowledge graph with multi-agent reinforcement learning. The model enabled learning state risk identification and personalized intervention. The study conducted a systematic evaluation using comprehensive learning behavior data from seven selected courses. The results indicated that the model achieved a high level of risk identification at an early stage. The recall values for all courses ranged from 0.879 to 0.896. As the learning process progressed, accuracy steadily increased to above 0.889. The F1-score remained between 0.842 and 0.871 across all stages, which demonstrated strong stability. Furthermore, the intervention strategies significantly improved learning trajectories across two experimental semesters. Students’ learning activities showed continuous improvement over time. Behavioral fluctuations and breakpoint frequency were both markedly reduced. These findings confirmed that the model consistently enhanced learning motivation, stabilized learning rhythms, and optimized patterns of resource utilization.
Multi‐Atom Sub‐Nanometer Assemblies on Interpenetrating Multi‐Chambered N/C Nanospheres
ABSTRACT The atomically dispersed catalysts have received much attention due to maximum atom utilization and enhanced catalytic performance. Compared with the widely studied single‐atom catalysts and dual‐atom catalysts, the multi‐atom catalysts (MACs) have unique features of collective effect of multiple metal atoms and more designable and tunable coordination environments that are desirable for catalytic reactions. Up to now, the research on MACs remains quite scarce, mainly restricted by the synthetic difficulty in precisely controlling the composition and arrangement of multiple metal atoms in MACs. Herein, we report a versatile soft–hard dual template route for the synthesis of both mononuclear MACs and heteronuclear MACs, which are stabilized on interpenetrating multi‐chambered N/C nanospheres. The as‐synthesized Ni/Cu‐MAC exhibits high performance for electrocatalytic CO 2 reduction reaction, delivering a CO Faraday efficiency of >99% at low required potentials (−0.26 V to −0.56 V). A cathode energy efficiency >75% is achieved at an industrial current density of 0.60 A cm −2 , representing highly competitive performance among the reported CO 2 ‐to‐CO electrocatalysts. The experimental and computational results demonstrate the synergistic effect between the atomically dispersed Ni and Cu for promoting the catalytic conversion of CO 2 to CO.
Investigation of early-age cracking experiments in concrete and research on improvement methods
A Sacrificial Seed Layer Strategy for Hierarchical MFI Zeolite Membranes With Enhanced Butane Isomer Separation Performance
ABSTRACT Zeolite membranes hold significant promise for industrially relevant gas separations; however, balancing high permeance and selectivity remains a persistent challenge. Although thinning the selective layer improves permeance, achieving precise fabrication remains a major bottleneck. Herein, we demonstrate a sacrificial seed layer strategy enabling the fabrication of hierarchical MFI zeolite membranes (HMFI) on commercial α‐Al 2 O 3 tubes. Through precise modulation of synergistic ionic interactions within the synthetic gel and the nutrient supply rate during crystallization, the seed layer functions as both a porogen to generate a macroporous sublayer and a “fertilizer” to promote surface gel crystallization, forming a dense top layer. The resulting HMFI membrane delivers more than a fourfold enhancement in the ideal selectivity of n ‐butane and isobutane ( n‐ / i‐ butane), accompanied by a 37% rise in n ‐butane permeance compared with conventional MFI membranes. Moreover, with a 10/90 n‐ / i‐ butane mixture, it exhibits an excellent separation factor of 158, along with a high n ‐butane permeance (188 × 10 −9 mol m −2 s −1 Pa −1 ), representing the highest performance for reported membranes on tubular supports. The strategy not only enables a novel route to high‐performance membranes, but also enriches the conceptual framework of the conventional secondary growth protocol, offering new avenues for microstructural engineering.
Information-theoretic grid topology reconstruction using low-precision smart meter data
Abstract Accurate knowledge of power grid topology is a prerequisite for effective state estimation and grid stability. While data-driven methods for topology reconstruction exist, the minimum requirements for measurement quality, specifically regarding quantization, precision, and sampling frequency, remain under-explored. This study investigates the data fidelity required to reconstruct distribution grid topologies using voltage magnitude measurements. Adopting an information-theoretic approach, we utilize the Chow–Liu algorithm to generate maximum spanning trees based on mutual information. Rather than proposing a new reconstruction algorithm, our primary contribution is a comprehensive sensitivity analysis of the measurement data itself. We systematically evaluate the impact of data bit-depth, significant digit truncation, time-window length, and different mutual information estimators on reconstruction accuracy. We validate this approach using IEEE test cases (via MATPOWER) and time-series data from GridLAB-D. Our results demonstrate that grid topology can be successfully recovered even with highly quantized 8-bit data or millivolt-level precision. However, performance degrades significantly when downsampling intervals exceed 20 min or when data availability is limited to short durations. These findings establish an optimistic theoretical lower bound, suggesting that costly high-precision instrumentation may not be strictly necessary for structural inference under ideal conditions. This rigorous baseline provides a foundation for future evaluations of noisy real world smart meter data and hybrid approaches that incorporate existing engineering priors.
Tailoring the Electric Double Layer for Advanced Rechargeable Batteries: Mechanisms, Strategies, and Outlook
ABSTRACT The structure of the electric double layer (EDL) at the electrode/electrolyte interface functions not only as the physical arena for electrochemical reactions, but also as the fundamental determinant of battery kinetics, interface stability, and cycle life. Although the solid electrolyte interphase (SEI)/cathode electrolyte interphase (CEI) film has been extensively studied, the microstructure and macroscopic performance of the EDL as precursor to interfacial film formation lacks systematical elucidation. This review aims to provide a comprehensive overview of the theoretical evolution of EDL models alongside their pivotal roles and regulation strategies in advanced batteries. We first retrace the development from classical Helmholtz models to modern microscopic theories. Subsequently, we delve into distinct regulation mechanisms and strategies tailoring EDL chemistry across various rechargeable battery systems such as Li‐based, Zn‐based and other metal‐ion batteries. These approaches encompass electrolyte optimization, electrode engineering, interface modification, and external field regulation, intending to suppress side reactions, guide uniform metal deposition, and stabilize electrode interfaces. Furthermore, advanced techniques for simulating and characterizing the microstructure of the EDL are discussed. Finally, the review outlines current challenges and forward‐looking perspectives on multidimensional rational design, data‐driven screening, operando characterizations, and applications under extreme conditions, providing theoretical guidelines for interface engineering of next‐generation batteries.
Combined nectar nonsugar compounds enhance bumblebee cognition
Technical and economic assessment of seven nanoparticle families for enhanced oil recovery in the Azadegan oil field
Abstract This study bridges the critical gap between laboratory promise and field-scale economic decision-making for nanoparticle-enhanced oil recovery (EOR) by developing an integrated, field-calibrated techno-economic framework specifically for the Azadegan Sarvak reservoir. We evaluate seven nanoparticle families-silica (SiO₂), nanoclay, magnetic Fe₃O₄, titania (TiO₂), graphene oxide (GO), zinc oxide (ZnO), and polymeric nanocapsules. Key innovations include: (1) a novel Laboratory-to-Field scaling procedure that translates core-flood metrics into field recovery estimates, (2) a risk-adjusted discounting method incorporating technical and environmental uncertainty, and (3) a quantitative Environmental Risk Index (ERI) to guide sustainable deployment. These components are integrated into a Monte Carlo–enhanced Discounted Cash Flow (DCF) model to provide a robust, transparent decision-support tool. The model incorporates a 10,000-run Monte Carlo uncertainty analysis, a risk-adjusted discounting method, and a quantitative Environmental Risk Index (ERI). Under a base oil price of 70 USD/bbl, SiO₂ and nanoclay emerge as the most economically attractive options (positive median NPV, IRR > 40%, median payback ≈ 3 years), benefiting from low cost and robust performance. TiO₂ and ZnO are conditionally economic, dependent on thermal stability and operational constraints. In contrast, GO and polymeric nanocapsules present high-risk/high-reward profiles at current unit prices, while Fe₃O₄ is uneconomic when the substantial CAPEX for magnetic guidance is included. Sensitivity analysis identifies oil price, incremental recovery (ΔRF), and nanoparticle unit price as the dominant value drivers. The study provides prioritized pilot recommendations, explicit technical/cost thresholds for scale-up, and a zonal implementation guide, delivering a practical decision-support tool for nano-EOR deployment in Azadegan.
Scalogram-CNN fusion with interactive 3D parameter exploration for inter-patient ECG classification
Green synthesis of gold nanoparticles utilizing Raphanus sativus root extract: characterization and in vitro investigation of its antioxidant, antimicrobial and anticancer impact
Abstract Gold nanoparticles (AuNPs) are recently investigated as substantial tools in cancer therapy due to their biocompatibility and tunable surface properties. Raphanus sativus or red radish root (RRR), a Cruciferaceae vegetable, has notable medicinal and nutritional value owing to its highly polyphenolic content and antioxidant property. The aim of this study is to synthesize the gold nanoparticles by a green method using the Raphanus sativus root extract and to study its antioxidant, antimicrobial and anticancer impact. The resultant RRR-AuNPs were characterized by TEM, UV/Vis spectroscopy, DLS, Zeta potential, XRD, AFM, BET and pore size analysis and assessed for antioxidant, antimicrobial and anticancer activity. The greenly synthesized gold nanoparticles showed ruby red color with a distinct absorption peak at 526 nm. The resultant AuNPs were spherical with an average size of 31 nm and average zeta potential of -36.8 mV. Results clearly demonstrated that the RRR-AuNPs possess high total polyphenolic content and potent antioxidant activity (DPPH). Moreover, RRR-AuNPs exhibited a considerable antibacterial efficacy against Staphylococcus aureus (80.9% inhibition) and Escherichia coli (78.22% inhibition) with a uniform MIC value of 0.977 µg/mL. Furthermore, RRR-AuNPs reduced the viability of Caco-2 and HepG2 cancer cells to 31.9% and 26.4%, respectively, at 250 µg/mL. Notably, a remarkable Antimicrobial Selectivity Index (SI) of 303.04 was established, confirming a toxicity over 300-fold higher toward bacterial pathogens than toward normal HEK-293 cells. These findings highlight RRR-AuNPs as a promising, biocompatible platform with potent antioxidant, antimicrobial, and anticancer properties for biomedical applications.
Experimental sepsis causes SERCA2 expression in white adipose tissue but not classical browning
Abstract Sepsis causes muscle wasting and cachexia but mechanisms remain unclear. Cachexia in cancer and burn injury is partly attributed to ‘browning’; where white adipose tissue (WAT) develops a catabolic, thermogenic brown adipose tissue-like phenotype. We hypothesised that sepsis-induced muscle wasting is caused by browning. 58 male Wistar rats were randomised to sham (n = 17) or experimental sepsis induced by intraperitoneal zymosan (n = 41). Tibialis anterior mass was measured on Days 3 and 14. Browning was sought using whole body and WAT respirometry, RNA-sequencing, immunoblot, thermal imaging and multi-photon microscopy of WAT. Fourteen-day mortality in rats receiving zymosan was 17%. In survivors, body mass loss peaked at day 3 and persisted to day 14 with associated tibialis anterior muscle mass loss. Zymosan peritonitis caused hypermetabolism during the late recovery phase (Days 11–14), but no difference in epididymal white adipose tissue temperature nor oxygen flux. At Day 14 transcriptomics showed inflammation but no increase in uncoupling protein (UCP)-1 at transcript or protein levels. SERCA2 protein was however increased fourfold in retroperitoneal WAT at day 14 ( p = 0.016). Rats recovering from zymosan peritonitis developed muscle wasting and cachexia associated with WAT inflammation and whole-body hypermetabolism. No evidence of browning was seen at functional, transcriptomic or protein levels, therefore our data do not support the hypothesis of classical browning as a driver of sepsis-induced muscle wasting and cachexia. SERCA2 protein expression was however increased in retroperitoneal WAT at day 14.
Seismic performance of a novel bracing system including cable-ring-arc together with MR damper in a steel frame
Anemia and risk of dementia in the general population: a propensity-score based cohort study
Study on the internal leakage pneumatic noise acoustic characteristics of spring-loaded full-lift safety valve
Piezoresistive and mechanical performance of nano-TiO₂-modified 3D-printed cementitious composites
TSG-6 protects orbital fibroblasts via anti-inflammation and anti-fibrosis effects in thyroid eye disease
A blockchain-enabled IoT framework for smart electro-medical waste management
Comparative evaluation of international and population-specific risk prediction models for type 2 diabetes: evidence from the kharameh cohort of the PERSIAN study
Hypertension and incidence of kidney stones: a prospective study based on the data of the Rafsanjan cohort study
A network analysis of personality traits, mentalizing, and psychological health in Chinese college students
Abstract The rising prevalence of anxiety and depression among college students constitutes a significant public mental health challenge. While personality traits (e.g., neuroticism) and impairments in mentalizing capacity are recognized as key vulnerability factors, their complex interplay in contributing to psychological distress remains inadequately elucidated. Network analysis offers a novel paradigm for visualizing this intricate system as an interconnected web of symptoms and traits. This study employed a network approach to investigate the interrelationships among personality traits, mentalizing, and psychological distress in a large sample of Chinese college students. The primary aims were to identify the most central (influential) elements within the network and, crucially, to detect bridge nodes that connect different psychological domains, thereby pinpointing potential targets for precise intervention. A cross-sectional survey was conducted among 5,140 Chinese undergraduates. Assessments included the Symptom Checklist-90 (SCL-90), Mentalizing Questionnaire (MZQ), Reflective Functioning Questionnaire-8 (RFQ-8), Eysenck Personality Questionnaire (EPQ), and University Personality Inventory (UPI). Regularized partial correlation networks were estimated using the LASSO-EBIC method (γ = 0.5), which inherently controls for multiple comparisons through regularization. Node centrality was indexed by expected influence (EI), and bridge centrality by bridge expected influence (bEI). Non-parametric bootstrap tests with 20,000 resamples were used to evaluate network accuracy and stability. Depressive symptoms (SCL-3) and neuroticism (EPQ-2) showed the highest centrality. Neuroticism (EPQ-2) was the strongest bridge node (bEI = 0.93), with the strongest edge in the network linking to impaired mentalizing (MZQ-1; weight = 0.474). Mean node predictability was 0.51. Network stability was excellent (CS coefficient = 0.75). Mentalizing constructs were represented by MZQ subscales, which also reflect the conceptual content of the RFQ-8: hypermentalizing (MZQ-2) corresponds to RFQ-C (certainty), and hypomentalizing (MZQ-3) corresponds to RFQ-U (uncertainty). Findings highlight robust concurrent associations between neuroticism, mentalizing impairment, and psychological distress in Chinese college students. Neuroticism is strongly associated with distress partly through its link to reduced mentalizing capacity. These cross-sectional results provide a framework for targeted preventive strategies in student populations, although causal inferences cannot be drawn from this observational design.