Tạp chí Khoa học - Trường Đại học Sư phạm Hà Nội: Khoa học Tự nhiên

ON CONVEX COMBINATIONS OF GRAZING LINEAR MODES FOR 2-DEGREE-OF-FREEDOM VIBRO-IMPACT SYSTEMS WITH 1:2 RESONANCE

Le Thi Huong and Stephane Junca — 2026-04-07

Grazing periodic orbits in vibro-impact mechanical systems are known to generate nonlinear normal modes. In the absence of internal resonance, many of the essential dynamical features already appear in two-degree-of-freedom (2-dof) mass–spring chains. In this note, we focus on such a system under the simplest internal resonance condition. Two grazing linear modes, denoted as GLOM1 and GLOM2, associated with the first and the second linear modes, play a key role.We will show that the grazing linear orbit generating one-impact-per-period (1IPP) solutions is neither GLOM1 nor GLOM2, but rather a convex combination of the two. This leads to the introduction of the set GLOc = [GLOM1, GLOM2] of grazing linear orbits. For all GLOc except the endpoint modes, the free-flight time to impact admits a particularly simple structure, providing a convenient framework for the future study of the nonlinear behavior of such grazing linear orbits.

GLOBAL ATTRACTOR FOR A CLASS OF REACTION-DIFFUSION SYSTEMS

Tran Thi Quynh Chi — 2026-04-07

We study the initial boundary value problem for a class of reaction-diffusion systems in bounded smooth domains, where one equation has a small diffusion coefficient δ > 0, and also the corresponding limit system is formally obtained when δ = 0. We prove the existence of global attractor Aδ for the dynamical system generated by the system in both cases δ > 0 and δ = 0. Moreover, the upper semicontinuity of the global attractor Aδ at δ = 0 is also investigated.

PROPER HOLOMORPHIC MAPS AND B−REGULAR DOMAINS IN Cn

Tran Duc Hieu — 2026-04-07

We begin by proving that B-regularity is preserved under proper holomorphic mappings of domains. We next investigate B-regularity in the setting of bounded Reinhardt domains in Cn. These results together demonstrate that B-regularity can be carried from well-understood Reinhardt model domains to more complicated domains via explicit proper holomorphic maps.

COLLECTIVE EXCITATIONS OF A BOSE-EINSTEIN CONDENSATE TRAPPED IN A TWO-CHANNEL POTENTIAL

Nguyen Viet Hung — 2026-04-07

In this paper, we investigate the collective excitations of an attractively interacting Bose–Einstein condensate confined in a two-channel potential. Using the Bogoliubov approximation, we derive the Bogoliubov–de Gennes equations and determine the excitation spectrum numerically. Two distinct modes are identified: the well-known Josephson mode, associated with periodic inter-channel tunneling, and a novel antisymmetric translational mode, describing out-of-phase oscillations of the condensate center of mass. Unlike conventional dipole oscillations, the latter arises from the interplay between inter-channel coupling and attractive nonlinearity and does not follow the Kohn theorem. We then employ the variational approximation to describe the condensate dynamics under these excitations and to extract the corresponding mode frequencies. The results are further validated by real-time simulations of the governing Gross–Pitaevskii equations, showing good agreement between analytical and numerical approaches.

MACHINE LEARNING ACCELERATED PROPERTY PREDICTION AND SCREENING OF STABLE LITHIUM-BASED BATTERY MATERIALS

To Anh Duc — 2026-04-07

Rapid screening of battery materials is critical to meeting the growing demand for high-performance energy-storage solutions. In this study, we employ a data-driven approach to evaluate and characterize stable lithium-containing compounds with potential applications in next-generation batteries. We retrieved a dataset of 1,846 stable lithium-based materials from the Materials Project database using the mp-api. By leveraging the matminer library, we generated a comprehensive set of composition-based and structure-based features, including elemental- property and density descriptors. A Random Forest Regressor was trained to predict the band gap of these materials, a key electronic property that determines their suitability as electrolytes or electrodes. The model achieved a mean coefficient of determination (????2) of 0.77 ± 0.04 and a mean absolute error (MAE) of 0.48 ± 0.05 eV across 10-fold cross-validation, demonstrating robust predictive capability. Featureimportance analysis revealed that electronegativity and Mendeleev number are the most significant predictors of band-gap energy. These findings underscore the effectiveness of machine learning in accelerating property prediction for battery materials and provide a pathway to more efficient experimental targeting.

RECOGNIZING AND UNDERSTANDING AMERICAN SIGN LANGUAGE USING DEEP LEARNING AND MEDIAPIPE

2026-04-07

Sign language serves as the primary medium of expression serves as the primary medium of expression for deaf and hard-of-hearing individuals. However, interpersonal interaction remains challenging, as they primarily rely on sign language to express their thoughts. To address this issue, this study proposes an automated sign language recognition and interpretation framework, integrating static and dynamic recognition components. Specifically, a Convolutional Neural Network (CNN) is employed for static gesture classification, while a hybrid CNN- Long Short-Term Memory (CNN-LSTM) architecture is utilized to capture the spatiotemporal features of dynamic signs. Furthermore, MediaPipe is leveraged for robust landmark localization to enhance feature extraction. The American Sign Language (ASL) dataset used in this research ensures diversity in sign representation, including variations in hand shapes, positions, and movements. The proposed models achieved high accuracy, with the CNN model reaching 93.1% and the CNN-LSTM model achieving 94.1% on test datasets, confirming their effectiveness in ASL recognition tasks.

NEW RECORDS AND CONSERVATION STATUS OF BATS (MAMMALIA: CHIROPTERA) IN LANG SON UNESCO GLOBAL GEOPARK, NORTHEASTERN VIETNAM

Nguyen Van Viet and Vu Dinh Thong — 2026-04-07

Lang Son UNESCO Global Geopark contains extensive karst landscapes, large cave systems, and diverse vegetation types that provide highly suitable habitats for bat species. This study updated the bat species composition of the geopark through nine field surveys conducted between February 2022 and July 2025 and through a review of relevant literature. A total of 26 species belonging to 14 genera and 7 families were recorded during the surveys, including five species first recorded for the geopark. When combined with previous records, the bat fauna of the geopark currently comprises at least 41 species representing 20 genera and 8 families. Of these, 12 species are listed in the 2024 Red Data Book of Vietnam, and six species are classified as Near Threatened, Vulnerable, or Endangered in the IUCN Red List of Threatened Species, indicating national and global conservation importance. In addition, the taxonomic status of nine species is updated in accordance with recent systematic revisions. These results provide a robust baseline for future research, long-term monitoring, and conservation planning, particularly for species of high conservation priority.

PREDICTING THE POPULATION TREND OF THE RHESUS MACAQUE (Macaca mulatta) AT SON TRA PENINSULA, DA NANG CITY

Nguyen Vinh Than, Nguyen Tuan Anh, Nguyen Dinh Duy and Nguyen Van Linh — 2026-04-07

The Rhesus macaque (Macaca mulatta) population at Son Tra Peninsula, Da Nang city, is experiencing conflicts with both residents and tourists, manifested in aggressive foraging behaviors and frequent incursions into residential areas. This study evaluated whether the rhesus macaque population on the Son Tra Peninsula has approached the carrying capacity of the habitat and assessed its projected trajectory under different snaring scenarios. In this study, a transect survey method was used to estimate the number of individuals in the macaque troops and the total number of individuals in the entire Son Tra Peninsula area. The age and sex structure of the troops was also recorded using this method. Population size, age–sex structure, group composition, and relevant ecological and demographic parameters were entered into Vortex 10 to simulate population trajectories over the next 50 to 100 years. The results show that under current conditions, with a total of 1487 individuals observed, the Rhesus macaque population in Son Tra has the potential to increase to approximately 1922 individuals. If the rate of illegal snaring consistently exceeds 10% of the total population at that time, the population size will begin to decline and gradually become extinct within 100 years.

HUMAN-MACAQUE INTERACTIONS ON SON TRA PENINSULA, DA NANG CITY: MANAGEMENT IMPLICATIONS FOR SUSTAINABLE COEXISTENCE

2026-04-07

This study examines the dynamics of human-macaque interactions on the Son Tra Peninsula, focusing on the influence of visitor and primate characteristics. Statistical analysis revealed that interaction distances predominantly ranged between 2 and 3 meters, suggesting a high degree of habituation within the population. A critical disparity was observed between visitor groups, as domestic visitors provisioned macaques with anthropogenic food significantly more frequently than international visitors (χ2 =22.123, p < 0.001). Among the primates, immature macaques were identified as the most active and bold age group, consistently displaying higher interaction intensity than adults or infants (F = 67.75, p < 0.05). Spatial analysis confirmed that Troop 8 and Troop 3 are the most habituated groups, often engaging in high-intensity proximity behaviors. These findings highlight the urgent need for demographic-specific management interventions to mitigate the risks of disease transmission and primate dependency on anthropogenic food.

SPECIES COMPOSITION AND DISTRIBUTION OF LEAFHOPPERS (HEMIPTERA: AUCHENORRHYNCHA: CICADELLIDAE) IN XUAN SON NATIONAL PARK, PHU THO PROVINCE

Nguyen Thi Thu Huong, Christopher H. Dietrich and Bui Minh Hong — 2026-04-07

The species composition and distribution characteristics of leafhoppers (Hemiptera: Cicadellidae) were investigated in Xuan Son National Park, Phu Tho Province. Direct surveys were conducted across three distinct habitats (natural forest, plantation forest, and grassland-shrub) over four sampling periods corresponding to the four seasons of the year. A total of 36 species belonging to 19 genera, 11 tribes, and 6 subfamilies were recorded. The abundance of leafhopper species in Xuan Son National Park fluctuated seasonally, decreasing from summer to winter and recovering in spring; nevertheless, species evenness remained consistently high and stable (J' > 0.80), suggesting a relatively undisturbed ecosystem. Species diversity also varied by habitat: natural forest exhibited the highest diversity (H’ = 2.95; S = 27 species, 75.0%), while plantation forest, despite fewer species, maintained high diversity and balance (H’ = 2.95; S = 22; 61.1%; J’ = 0.95). In contrast, the grasslandshrub habitat showed the lowest diversity and significant imbalance (H’ = 2.36; S = 20; 55.6%; J’ = 0.79). However, species composition across the three habitats was highly similar. Representative species were identified as follows: 11 species were recorded in all three habitats, 11 in two habitats, and 14 were unique to a single habitat.

ANTIOXIDANT ACTIVITY AND SELECTED METABOLIC EFFECTS OF ETHANOL EXTRACT FROM MANGOSTEEN (Garcinia mangostana L.) PERICARP IN HIGH-FAT DIET–INDUCED MICE: AN IN VITRO AND IN VIVO STUDY

Nguyen Thi Hien, Le Thi Tuoi, Tran Thu Ngan and Nguyen Thi Hong Hanh — 2026-04-07

Mangosteen (Garcinia mangostana L.) pericarp is a rich source of polyphenolic xanthones and other bioactive compounds with potential antioxidant and metabolic regulatory properties. This study investigated the in vitro antioxidant activity and selected in vivo metabolic effects of ethanol extract from mangosteen pericarp. Mangosteen pericarp collected from Lam Dong province was extracted with 60% ethanol. The extract was characterized for total phenolic content by the Folin–Ciocalteu method. Antioxidant capacity was evaluated using the DPPH radical scavenging assay. For the in vivo study, male white mice (Mus musculus) were divided into four groups: a regular diet (RD), a high-fat diet (HFD), an HFD with extract at 300 mg/kg, and an HFD with extract at 400 mg/kg. Liver weight, spleen weight, and fasting blood glucose levels were measured after 8 weeks. The extract exhibited a high total phenolic content (200.16 ± 17.71 mg gallic acid equivalents/g extract) and showed dose-dependent DPPH radical scavenging activity with a half maximal inhibitory concentration (IC50) value of 118.87 ± 13.56 μg/mL, indicating a moderate but reproducible antioxidant capacity. In vivo, the HFD group showed significantly increased blood glucose levels (10.38 ± 3.54 mmol/L) compared to RD controls (7.57 ± 3.45 mmol/L). Treatment with the extract at 400 mg/kg significantly reduced glucose levels (5.16 ± 0.93 mmol/L, p < 0.05) compared to the HFD group. The extract at 300 mg/kg showed intermediate effects (9.89 ± 2.77 mmol/L). No statistically significant differences were observed in liver or spleen weights among the groups. In summary, the mangosteen pericarp ethanol extract is rich in phenolics, exhibits moderate antioxidant activity in vitro, and demonstrates dose-dependent protective effects against HFD-induced hyperglycemia. While the in vivo metabolic assessment was limited to selected endpoints, these findings provide preliminary evidence supporting the potential of the mangosteen pericarp extract for further investigation in diet-induced metabolic dysfunction models.

RAINFALL VARIABILITY AND EXTREME PRECIPITATION EVENTS UNDER CLIMATE CHANGE: A CASE STUDY IN PHU THO PROVINCE, VIETNAM

Dao Ngoc Hung, Tran My Hang, Cu Thi Phuong, Trinh Thi Diu and Nguyen Khac Anh — 2026-04-07

This study investigates rainfall variability in Phu Tho province, Vietnam, focusing on total annual rainfall, the frequency of heavy rainfall days (≥ 50 mm/day), and very heavy rainfall days (≥ 100 mm/day), alongside rainy season characteristics using daily rainfall data from 1995 to 2024. Trends and variability in rainfall were analyzed using statistical methods, including the Mann–Kendall test and Sen’s slope estimator. The results indicate that trends in total annual rainfall are generally weak and spatio-temporal heterogeneity. In contrast, distinct climate change signals are evident in the intensification of rainfall and extreme weather patterns. Heavy and very heavy rainfall events contribute a disproportionately large share of total annual rainfall, particularly in mountainous terrain, despite their relatively low frequency. Furthermore, changes in rainy season duration remain limited, suggesting that recent climate impacts are primarily manifested through increased rainfall intensity rather than shifts in seasonality. These findings underscore the escalating significance of extreme rainfall events for disaster risk management and climate change adaptation in Phu Tho province.

SPATIOTEMPORAL CHARACTERISTICS AND TRENDS OF HEAT EXTREMES IN LAO CAI PROVINCE, VIETNAM

Dao Ngoc Hung, Tran Bich Diep, Ly Thuy Nga, Bui Thi Hong Tuyet and Cu Thi Phuong — 2026-04-07

Heat extremes are among the most significant manifestations of climate change, posing increasing risks to human health, agriculture, and socio-economic activities. This study analyzes the characteristics and long-term trends of heat extremes in the Lao Cai region, northern Vietnam, over the period 1975 - 2024 using daily maximum temperature data from meteorological observation stations. The occurrences of hot days and heat waves were identified based on thresholdbased indices and consecutive-day criteria. The results reveal a clear intensification of heat extremes, with significant increases in the frequency, intensity, and duration of hot days and heat waves, particularly in low-elevation areas. In contrast, stations located at elevations above 900 m experience minimum hot days, indicating a strong influence of topography on the spatial distribution of thermal extremes. Heat extremes exhibit strong seasonality, occurring mainly during the warm season from May to September. These findings demonstrate that mountainous regions such as Lao Cai are increasingly affected by extreme heat and provide a crucial scientific basis for heat risk assessment and climate adaptation planning.

RETRIEVAL OF PM2.5 CONCENTRATIONS USING MODIS-DERIVED AEROSOL OPTICAL DEPTH IN HO CHI MINH CITY, VIETNAM

Nguyen Van Manh, Nguyen Dai Long, Hoang Thi Thu Ha and Luong Thi Kim Cuc — 2026-04-07

Air pollution characterized by fine particulate matter (PM2.5) has emerged as a critical environmental problem in Vietnam, particularly within densely populated metropolitan areas like Ho Chi Minh city. The integration of remote sensing (RS) and geographical information systems (GIS) offers a robust framework for monitoring PM2.5 due to its extensive spatial coverage, high temporal resolution, and cost-effectiveness. This study estimates ground-level PM2.5 concentrations by establishing a correlation between in-situ measurements from six monitoring stations and Aerosol Optical Depth (AOD) values retrieved from MODIS satellite imagery. The spatial, temporal, and seasonal dynamics of PM2.5 in Ho Chi Minh city during the 2020 – 2025 period. Results indicate that the highest concentrations are concentrated in central urban districts, averaging 24.0 μg/m³. During the rainy season, the average of PM2.5 concentration drops to 14.4 μg/m³. This seasonal reduction is primarily attributed to enhanced atmospheric dispersion and the wet deposition (rainwash). Overall, the study demonstrates that utilizing satellite-derived AOD provides a more comprehensive spatial understanding of PM2.5 distribution, thereby supporting evidence-based air quality management and sustainable urban planning.