Browse Articles

A hybrid KF–GAIN framework for dynamic state estimation in islanded microgrids under missing PMU data and line impedance uncertainty

Abstract The similar time constants of voltage source inverter-based distributed generation units and other components in islanded microgrids (IMGs) pose significant challenges for dynamic state estimation (DSE). To address these issues, particularly under conditions of missing measurement data and uncertainties in line impedances, this paper proposes a linear hybrid framework that combines the Kalman Filter (KF) with a deep learning model called the generative adversarial imputation network (GAIN) for DSE in IMGs. GAIN robustly imputes missing data while preserving the statistical properties of the original dataset, aided by a hint mechanism that guides the generator to produce values closely following the true data distribution. The proposed KF-GAIN model is evaluated on an islanded IEEE 33-bus test system equipped with distributed generation units and phasor measurement units (PMUs), incorporating temperature-dependent line impedance variations modeled via Monte Carlo simulation. The IMG is first simulated in MATLAB/Simulink; subsequently, the measurement data are then transferred to Python for missing data imputation using GAIN, followed by linear DSE via the KF. To assess GAIN’s performance, autoencoder (AE) and variational autoencoder (VAE) models are employed as baseline methods. Hyperparameters (GAIN, AE, VAE) were selected by grid search; robustness was assessed via sensitivity to single and pairwise PMU outputs. GAIN achieves a significantly lower imputation error (0.02705) compared to AE (0.35872) and VAE (0.20344), demonstrating superior accuracy. Furthermore, the proposed framework maintains high estimation accuracy, with average errors of 0.46% and 0.72% for 20% and 40% missing data, respectively, across multiple scenarios.

Understanding the value of a forecast using an online game

This research illustrates the use of an online game to study the value of forecasts under conditions of information uncertainty. The objective of the game is for players to plant crops in field locations in a way that maximizes a payoff, with the option of paying for forecasts that help make better decisions. We found that while players fall short of theoretically perfect play, they generally make decisions that improve their score. Players both under and overpay for forecasts that reduce uncertainty, but exhibit behaviour that shows they roughly understand the expected value of forecasts. The small sample size and demographic homogeneity of the participants limit our ability to generalize these results, but the results suggest that game environments can be useful platforms for augmenting existing primary data collection methods. This is especially poignant today since with the assistance of AI, researchers with little to no programming experience can now design and develop games in ways that would have been prohibitively costly less than a decade ago.

Experience and personality modulate pupillary responses during real-time processing of within-language accent shifts

Abstract Speech perception shows substantial individual differences, but the mechanisms underlying this variability during real-time accent adaptation in natural conversation remain poorly understood. We used pupillometry to examine the temporal dynamics of processing effort while German listeners followed a semi-natural dialogue alternating between Standard German and an Alemannic regional variety of German. Growth curve analysis of the within-trial time course revealed switching costs and initial asymmetries, with greater effort for the regional variety and for switches from standard to regional accent. An analysis across the course of the dialogue further showed that the effect of accent driving these asymmetries diminished as listeners adapted to the regional variety. In contrast, switching costs remained stable across the dialogue, consistent with sustained attentional demands. Individual differences modulated these effects. A k-means cluster analysis yielded two clusters of listeners. The group characterized by greater exposure to Standard German, younger age, lower openness, and more favorable comprehensibility and pleasantness ratings for the standard speaker showed higher switching asymmetry, while overall switching costs did not differ reliably between profiles. These results are consistent with the interpretation that switching cost and switching asymmetry may reflect partly distinct underlying processes and that individual differences in language experience and personality selectively shape asymmetries during adaptation to within-first-language accent variation in real-time conversation.

Network analysis of smartphone addiction and sleep disorder symptoms in Chinese college students

Objective This study aims to examine the comorbid relationship between smartphone addiction and sleep disorders in Chinese college students. By constructing a comorbidity network, identifying core and bridge symptoms, and exploring potential directional associations among symptoms, this research intends to establish a theoretical foundation for targeted intervention strategies. Methods A total of 1842 Chinese college students were recruited through convenience sampling. The smartphone addiction and sleep disorder symptoms were assessed using the Smartphone Addiction Scale-Short Version (SAS-SV) and the Pittsburgh Sleep Quality Index (PSQI), respectively. The data analysis was conducted in three steps. First, an undirected comorbidity network was constructed using the Gaussian Graphical Model (GGM) to identify core and bridge symptoms. Second, a Bayesian network approach was employed to generate Directed Acyclic Graphs (DAGs) that explored potential directional associations among symptoms. Finally, network comparison tests and community detection analyses were performed to examine gender differences in the comorbidity network structure. Results The GGM comorbidity network exhibited a connection density of 0.80 and a global strength of 9.39. Within this network, PSQI2 (sleep latency), SA2 (difficulty concentrating), and SA5 (impatience without phone) were identified as core symptoms. PSQI2 (sleep latency), PSQI1 (subjective sleep quality), and SA9 (longer use than intended) were identified as bridge symptoms. Further analysis using the DAGs suggested statistical directionality from sleep disorder symptoms toward smartphone addiction symptoms. Notably, SA5 (impatience without phone) served as an initial node in the DAGs. Finally, network comparison tests indicated no significant differences in the GGM network structure between genders; however, distinct gender differences were observed in the community clustering patterns of symptoms. Conclusion In college students, smartphone addiction and sleep disorder symptoms interact to form a structurally stable comorbidity network. Consequently, interventions targeting core symptoms, bridge symptoms, and initial node could effectively interrupt the maintenance of this comorbidity.

Research on seepage and stability of embankment slopes under coupled rainfall-vehicle loading

Laboratory Inventory Management Engine (LIME): A free tool for managing laboratory inventories via barcode scanning and automatic cloud-based spreadsheet integration

Efficient inventory management remains a persistent challenge in research settings, where productivity and budget control are tightly linked to accurate tracking of consumables, reagents, and equipment for decision-making and successful research project execution. Many laboratories still rely on traditional methods, including handwritten logs or static spreadsheets, which are prone to human error and can result in stockouts, over-ordering, and inaccurate forecasting. Moreover, existing inventory systems often fall short in balancing affordability, convenience, and functionality. Here, the Laboratory Inventory Management Engine (LIME) is introduced as a system designed to bridge these deficiencies by leveraging accessible tools like in-the-cloud spreadsheet files and mobile devices. LIME uses a smartphone app to enable real time updates to the inventory log by scanning consumables QR codes and barcodes. The system allows inputting new items to be classified in predetermined catalog lists (e.g., 4º fridge, chemicals, antibodies, etc.). This dual smartphone and spreadsheet inventory prototype is anticipated to improve the operational efficiency of small academic research laboratories and early-stage startups to benefit their bottom lines.

Research on the evolution and driving characteristics of cultural-tourism integration efficiency in urban clusters of the Yangtze River Delta

Inhibition of neutrophil infiltration and NETs formation ameliorates neuropsychiatric and renal dysfunction in MRL/lpr mice with lupus

Background Patients with systemic lupus erythematosus often experience kidney and nervous system damage. It has been discovered that neutrophils and NETs play a significant role in lupus-related damage. However, whether inhibiting the infiltration of neutrophils into tissues can alleviate lupus-related brain and kidney damage remains unclear. Methods Lupus-prone MRL/lpr mice were assigned to four groups: MRL/MpJ control, MRL/lpr model, MRL/lpr + Avacopan, and MRL/lpr + Naringenin. Cognitive functions such as memory, emotion, and learning in the mice were assessed through open field tests, Y-maze tests, and water maze tests. Changes in kidney function were also evaluated in each group. Pathological changes in the brain and kidneys of the mice, as well as the infiltration of neutrophils and the expression of NETs-related markers NE and MPO, were observed using histopathological staining, immunohistochemistry, silver staining, and multiplex immunofluorescence staining. Additionally, changes in the expression of key inflammatory factors in the brain and kidneys were examined. Results Avacopan can improve cognitive function impairment and the progression of kidney damage in lupus-prone mice. Additionally, Avacopan has been found to reduce the infiltration of CD11b-positive and CD11b/CD16 double-positive cells into the brain and kidneys, as well as to decrease the expression of NETs-related markers NE and MPO. Notably, we observed that the main NETs marker in the kidneys of lupus mice is NE. Key inflammatory factors associated with lupus, such as IL-6, IL-17, and TNF-α, are found to be elevated in both the brain and kidney tissues of lupus mice, and Avacopan can reduce the expression of these inflammatory factors. Conclusion Neutrophils play a significant role in systemic lupus erythematosus-related brain and kidney damage. Inhibiting the infiltration of neutrophils can mitigate the inflammatory damage and NETs-related damage caused by their infiltration. However, the detailed mechanisms by which neutrophils cause tissue damage still need further clarification. For example, it remains to be explained why there is such substantial infiltration of neutrophils into tissues, what the cell signaling mechanisms involved in damage after neutrophil infiltration are, and whether there is any impact on the regulation of other immune cells such as macrophages and T lymphocytes. Further research is needed on these aspects.

Mechanical and durability performance of gypseous soils stabilized with zeolite and lime

HIV incidence among sexually active young males and females in Kisumu County, Western Kenya

Introduction HIV prevalence in Kisumu County, Kenya, is over four times the national average. We estimated HIV incidence among adults with multiple sexual partners and identified risk factors associated with HIV acquisition. Methods We enrolled adults aged 18–35 years reporting ≥2 sexual partners in the prior 3 months, recruiting primarily from villages with fisherfolk. HIV counseling and testing were performed every 3 months for up to 24 months. Demographic, health, sexual, and behavior questionnaires were administered every 6 months. Enrollment characteristics were compared between participants who did and did not acquire HIV using chi-squared tests. Bivariable Cox proportional hazards models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for potential risk factors. Results Between January 2017 and August 2021, 619 participants without HIV were enrolled, of whom 45% were female, 47% were ≥25 years old, 66% were single or widowed, and 46% had some secondary education. Eleven HIV seroconversions occurred over 1117.9 person-years (PY; incidence 9.84/1000 PY [95% CI: 5.22–17.03]), including eight in the first year of follow-up (13.84/1000 PY). Risk factors in unadjusted analyses included ≥4 sexual partners (HR 4.39 [95% CI 1.10-17.55]) and forced sexual activity (HR 9.15 [95% CI 1.17-71.51]). Conclusion HIV incidence was lower than expected, possibly reflecting prevention efforts or COVID-19-related behavior changes. Lower incidence in later follow-up may indicate a cohort effect from regular counseling. Increasing ART coverage and pre-exposure prophylaxis may have decreased HIV incidence in fishing communities of Western Kenya.

Optimizing adult nutrition and mating duration to improve mass rearing of Chrysoperla zastrowi sillemi

Antimicrobial resistance in Gram-negative bacteria from retail meat: Prevalence and public health implications

Retail meat is an important source of nutrition, but it may also serve as a reservoir of antimicrobial-resistant Gram-negative bacteria, posing a significant public health concern. This study investigated 123 retail meat samples collected from outlets in Sindh, Pakistan, including locally produced raw ground meat, imported raw ground meat, raw beef burgers, frozen chicken portions, and swabs from the outer surfaces of chicken carcasses, to identify Gram-negative bacteria and assess their antimicrobial susceptibility patterns. Gram-negative bacteria were recovered from all samples. Among the isolates, Escherichia coli was the most prevalent (37.4%), followed by Klebsiella spp. (25.2%), Salmonella spp. (17.9%), Enterobacter spp. (13.8%), Pseudomonas spp. (2.4%), Citrobacter spp. (2.4%), and Aeromonas spp. (0.8%). Identification was based on culture characteristics, Gram staining, biochemical testing, and API 20E confirmation. Antimicrobial susceptibility testing showed high levels of resistance among the isolates, particularly to ampicillin, amoxicillin-clavulanate, cefotaxime, ceftazidime, and ceftriaxone. E. coli , Klebsiella spp., Salmonella spp., and Enterobacter spp. exhibited substantial multidrug resistance, while comparatively lower resistance was observed for cefepime, gentamicin, and piperacillin-tazobactam in some species. Overall, resistance to multiple antimicrobial classes was common, indicating the widespread presence of multidrug-resistant Gram-negative bacteria in retail meat. These findings highlight the potential role of retail meat as a source of antimicrobial-resistant bacteria in the food chain and underscore the need for strengthened surveillance, improved hygiene practices, and antimicrobial stewardship in food animal production. A One Health approach is essential to help limit the spread of antimicrobial resistance across human, animal, and environmental sectors.

A randomized controlled trial of adjunctive speleotherapy in asthma, COPD and long COVID

Abstract Speleotherapy (underground climate therapy) is a non-pharmacological intervention for chronic respiratory diseases. This randomized controlled trial investigated whether a 3-week speleotherapy course (6 sessions, 2 h/week) improves respiratory outcomes in patients on standard background therapy with asthma, COPD, or Long COVID, and whether it affects blood CO 2 levels. The control group did not receive speleotherapy but continued their standard therapy. A total of 208 patients (asthma: n = 107; COPD: n = 59; Long COVID: n = 42) were enrolled across nine centers in Germany, Austria, and Italy. Assessments were conducted pre-intervention (T1), post-intervention (T2), and at 3-month follow-up (T3). Outcome measures included airway inflammation (FeNO), pulmonary function parameters (FVC% predicted values, FEV₁% predicted values, FEV₁/FVC, PEF% predicted values), and respiratory muscle strength (MIP and MEP in absolute values). In addition the following validated questionnaires were administered: Asthma Control Test (ACT), Asthma Quality of Life Questionnaire (AQLQ), COPD Assessment Test (CAT), St. George’s Respiratory Questionnaire (SGRQ), Nijmegen Questionnaire (NQ), and Fatigue Assessment Scale (FAS), along with the Long COVID questionnaire from the Median Clinic Group. CO 2 levels were assessed via capillary blood (SpCO 2 ) and end-tidal CO 2 (PetCO 2 ). Between-group comparisons used the Mann–Whitney U test; within-group changes were assessed with the Wilcoxon signed-rank test (Bonferroni-Holm corrected). In patients with asthma, the predefined primary endpoint (FeNO) showed no significant improvement. In contrast, patient-reported outcomes improved significantly, with clinically relevant gains in asthma control (ACT: p < 0.001) and asthma-related quality of life (total AQLQ: p = 0.005). Lung function parameters showed statistically significant but modest improvements at T2 (FVC: p = 0.011; PEF: p = 0.010; FEV₁ in participants < 70 years: p = 0.035). In patients with COPD, symptom burden improved according to CAT scores (p = 0.036), while no improvements in lung function were observed. Patients with Long COVID reported significant improvements in dysfunctional breathing (NQ: T2: p = 0.014), dyspnea (T2: p = 0.026; T3: p = 0.001), and “problems with stair climbing/muscle exertion” (T2: p = 0.042), as well as improvements in anxiety and sleep-related symptoms (T2: p = 0.021). No improvements in lung function were observed in this group. In the total cohort, the intervention group showed statistically significant improvements compared to controls in respiratory muscle strength (MIP: p = 0.002; MEP: p = 0.018) and dysfunctional breathing (NQ scores at T2: p = 0.007; T3: p = 0.017). In CO₂-rich speleotherapy centers, both SpCO₂ (p = 0.026) and PetCO₂ (p < 0.001) increased at T2. While speleotherapy did not improve FeNO, it was associated with clinically relevant improvements in patient-reported outcomes across disease groups. Changes in lung function and respiratory muscle strength were statistically significant but modest and should be interpreted with caution. Overall, speleotherapy may have direct effects on the airways and breathing regulation, with more consistent evidence for improvements in breathing patterns than for direct effects on the airways. Trial registration : DRKS00033365 (retrospectively registered).

RPINet: Dual-branch attention-guided fusion of remote sensing images and point clouds for urban scene segmentation

Semantic segmentation of large-scale urban point clouds is a fundamental yet challenging task due to the complex spatial structures, massive data volume, and irregular distribution of points. Existing methods typically rely solely on 3D geometry or naively fuse 2D and 3D features, leading to limited performance in capturing both fine-grained details and global semantic consistency. In this paper, we propose RPINet, a novel Remote-Projection and Intelligent Network that effectively integrates 2D remote sensing images with 3D point cloud data for enhanced semantic segmentation. RPINet adopts a dual-branch architecture where the 3D point branch extracts spatial features using PointNet++, transformers, and graph convolutional networks, while the 2D image branch leverages a Gaussian splatting projection and CNN-based encoding to retain texture and contour information. A hybrid attention-based fusion module dynamically weights intra-modal and inter-modal dependencies, enabling deep semantic interaction between modalities. In addition, we introduce an adaptive sampling strategy and a multi-objective loss function to optimize segmentation accuracy and geometric consistency. Extensive experiments on the challenging SensatUrban dataset demonstrate that RPINet achieves state-of-the-art performance with a mean IoU of 66.5%, outperforming existing methods by a significant margin. Our model also shows strong generalization ability on unseen datasets, confirming its robustness and practical applicability to real-world urban scenarios.

Hybrid BOTDA-DIC sensing for flexural behavior characterization of eco-concrete reinforced beams under four-point loading

Abstract This study investigates the performance of Brillouin Optical Time Domain Analysis (BOTDA) and Digital Image Correlation (DIC) for monitoring the flexural behavior of two reinforced concrete (RC) beams constructed using eco-concrete containing Palm Oil Boiler Ash (POBA) and nano-silica as partial cement replacement. Two simply supported beams (300 cm × 15 cm × 25 cm) with different concrete compressive strengths were tested under four-point bending until failure to evaluate the sensing performance under varying structural responses. Distributed strain from BOTDA and full-field displacement from DIC were compared with mid-span deflection measured by Linear Variable Differential Transformers (LVDTs). Both sensing systems successfully captured strain development and stiffness degradation. BOTDA recorded the progressive increase of tensile strain at the bottom fiber, reaching 10,000–18,000 µε, as cracking advanced, while DIC identified multiple flexural cracks with openings ranging from 0.05 to 3.28 mm. BOTDA-derived deflection showed good agreement with LVDT and DIC, with differences less than 7%. The findings highlight the complementary advantages of both systems: BOTDA provides distributed and continuous strain monitoring for global behavior, while DIC captures localized deformation and crack development with high precision. Together, these techniques offer a complementary and reliable approach for detailed structural assessment of reinforced concrete beams.

Integrative transcriptomic analysis identifies immune-associated candidate genes and altered immune cell infiltration in pulmonary arterial hypertension

Background Pulmonary arterial hypertension (PAH) is a progressive vascular disease characterized by immune dysregulation and pulmonary vascular remodeling. This study aimed to identify immune-associated hub genes in PAH using an integrative bioinformatics framework and to validate key candidates in an experimental model. Methods Three PAH lung transcriptomic datasets from the Gene Expression Omnibus (GEO) database were analyzed. Immune cell infiltration was estimated using single-sample gene set enrichment analysis (ssGSEA). Differentially expressed genes (DEGs) were identified and integrated through weighted gene co-expression network analysis (WGCNA) and protein-protein interaction (PPI) network construction. Hub genes were prioritized using multiple machine learning algorithms. A PAH-relevant murine model (Su5416 combined with hypoxia) was used for in-vivo validation by quantitative real-time PCR. Results A total of 8 hub genes were identified through integrative screening across multiple algorithms and were validated in independent datasets. Among these hub genes, BCLAF1 demonstrated the highest diagnostic performance. Immune infiltration analysis revealed significant alterations in T helper cell subsets in PAH. Correlation analysis indicated associations between hub genes and specific immune signatures, including positive correlations of CDC5L and RBM39 with Tgd cells, a negative correlation of ASH1L with neutrophils, and inverse associations of CTNNB1 and SMARCA5 with dendritic cells (DCs) and central memory T cell (Tcm) signatures. In the PAH murine model, BCLAF1, CDC5L, SMARCA5, and ASH1L were significantly upregulated in lung tissues, accompanied by enhanced collagen deposition. Conclusion This study identified BCLAF1, CDC5L, SMARCA5, and ASH1L as immune-associated hub genes in PAH and proposed a transcriptomic gene–immune prioritization framework. These candidates warrant further mechanistic investigation for their potential roles in PAH pathogenesis.

Evaluation of construction progress of smart highway: a Bayesian network model

Burkholderia cepacia complex (Bcc) in goats: First report in Bangladesh

The Burkholderia cepacia complex (Bcc) comprises a diverse group of opportunistic pathogens that are critically relevant to human and animal health. Bcc induces life-threatening infections of the lung in human patients with cystic fibrosis and chronic granulomatous disease and causes mastitis and abscesses in sheep and goats. Therefore, this research aimed to identify the presence of Burkholderia cepacia complex (Bcc) in goats in Bangladesh using serological tests such as the Glanders Rapid Detection Test Kit (GRDTK) and Enzyme-linked immunosorbent assay (ELISA) and molecular tests such as PCR, Sanger sequencing, and phylogenetic analyses. A total of 40 goat blood samples were collected, comprising 30 samples from goats exhibiting a history of abortion accompanied by respiratory symptoms and 10 samples from clinically healthy control goats. Among the 30 samples from symptomatic goats, eight isolates (26.67%) were observed to grow in Luria-Bertani broth, while there was no growth from the control group. Five (62.50%) were found positive for bacterial presence when analyzed using the 16S rRNA gene-specific PCR assay for broth-positive samples. In the Genomix GRDTK and ELISA tests, six isolates were identified as positive for Burkholderia spp. Of the eight broth culture-positive samples, two (25%) were found to be positive by genus-specific PCR using groEL primers and species-specific amplification using zmpA primers. Phylogenetic analysis of positive goat samples confirmed the presence of the Bcc, which showed 100% similarity with the strains from India, Japan, and China. The discovery is the first detection of Bcc in animals in Bangladesh. It raises significant concerns for animal and public health. These findings highlight the critical need for strengthened diagnostic strategies, improved microbiological surveillance, and further research into Bcc’s epidemiology and zoonotic potential within the One Health framework to understand better and mitigate its zoonotic risks.

The prevalence of E198A and F200Y single nucleotide polymorphism in Haemonchus contortus populations from Polish goat herds

Research on a financial fraud identification model by fusing a convolutional neural network

To address the challenges of low accuracy and poor real-time performance in existing financial fraud identification methods for listed companies, this paper proposes a hybrid identification model (CNN-SVM) that fuses a convolutional neural network with a support vector machine. Utilizing a comprehensive dataset of 7,429 samples from non-financial A-share listed companies penalized for fraud between 2007 and 2022, the study employs random oversampling to rebalance the minority class, resulting in a robust training set of 13,540 samples. The model leverages a CNN architecture to automatically extract high-level features from 87 indicators spanning corporate governance, financial oversight, and operational metrics, which are then classified using a linear-kernel SVM. Experimental results demonstrate that the CNN-SVM model achieves a qualitative leap in performance, yielding an AUC of 0.97, a recall of 0.99, and an F1-score of 0.97, significantly outperforming traditional logistic regression and random forest benchmarks. These findings suggest that the proposed framework effectively balances training efficiency with high precision, providing a superior tool for real-time financial risk control.