Search Results (32,013)

Lower limb exoskeleton systems require accurate recognition of the wearer’s movement intentions prior to action execution in order to achieve natural and smooth human–machine interaction. Surface electromyography (sEMG) signals can reflect neural activation of muscles before movement onset, making them a key physiological source for movement intention recognition. To improve sEMG-based recognition performance, this study proposes an innovative deep learning framework, IFIANet. First, a CNN–TCN-based spatiotemporal feature learning network is constructed, which efficiently models and represents multi-scale temporal–frequency features while effectively reducing model parameter complexity. Second, an IFIA (Frequency-Informed Integration Attention) module is designed to incorporate global frequency information, compensating for frequency components potentially lost during time–frequency transformations, thereby enhancing the discriminability and robustness of temporal–frequency features. Extensive ablation and comparative experiments on the publicly available MyPredict1 dataset demonstrate that the proposed framework maintains stable performance across different prediction times and achieves over 82% average recognition accuracy in within-experiments involving nine participants. The results indicate that IFIANet effectively fuses local temporal–frequency features with global frequency priors, providing an efficient and reliable approach for sEMG-based movement intention recognition and intelligent control of exoskeleton systems. Full article

Long-term partial discharge (PD) erosion is an important factor causing insulation failure. With the rapid development of power electronics and the widespread application of inverters, more insulation is subjected to repetitive frequency square wave signals. To understand the void PD characteristics of insulation subjected to repetitive frequency square wave signals, a finite element simulation model of void PD is established based on the ABC model. A time-domain waveform and a phase-resolved partial discharge (PRPD) plot of void PD under repetitive frequency square wave with different overshoot rates are simulated. Then, void PD is measured under different overshoot rates and compared with the simulation results. Finally, the influence of overshoot voltage on internal discharge is analyzed. The simulation and experimental results show that the overshoot rate positively correlates with statistical characteristics, such as the average PD number and maximum PD quantity. Void PD events mainly occur in the overshoot portion of the repetitive frequency square wave. Therefore, the overshoot portion of the repetitive frequency square wave is one of the key factors contributing to severe PD in insulation. Full article

Climate change is increasingly threatening the sustainability of coffee farming in northern Peru, particularly in the Amazonas region, where coffee cooperatives serve as vital socioeconomic hubs for thousands of families. This study analyzed historical climate data from 1979 to 2024 to project trends up to 2030, integrating local perceptions from coffee producers to identify trends, anomalies, and future scenarios within four coffee cooperatives in northern Peru. We examined variables such as precipitation, temperature, evapotranspiration, and wind speed using nonparametric statistical analyses and SARIMA time-series models. The findings indicate a steady increase in maximum and average temperatures, alongside greater irregularity in precipitation. Specifically, the Bagua Grande and COOPARM cooperatives are experiencing precipitation deficits, while the Alta Montaña and Ocumal cooperatives are facing excess rainfall. Additionally, we project an increase in evapotranspiration by 2030. Surveys conducted with coffee growers reveal a consensus regarding irregular rainfall patterns; however, there is less recognition of the rising temperature trends. This discrepancy emphasizes the importance of combining scientific data with local knowledge to develop more effective adaptation strategies at the cooperative level. We conclude that enhancing climate training and cooperative management is essential for improving the resilience of regional coffee farming. Full article

Background/Objectives: Histopathological images are fundamental for the morphological diagnosis and subtyping of lung cancer. However, their high resolution, color diversity, and structural complexity make automated segmentation highly challenging. This study aims to address these challenges by developing a novel hybrid metaheuristic approach for multilevel image thresholding to enhance segmentation accuracy and computational efficiency. Methods: An adaptive hybrid metaheuristic algorithm, termed SCSOWOA, is proposed by integrating the Sand Cat Swarm Optimization (SCSO) algorithm with the Whale Optimization Algorithm (WOA). The algorithm combines the exploration capacity of SCSO with the exploitation strength of WOA in a sequential and adaptive manner. The model was evaluated on histopathological images of lung cancer from the LC25000 dataset with threshold levels ranging from 2 to 12, using PSNR, SSIM, and FSIM as performance metrics. Results: The proposed algorithm achieved stable and high-quality segmentation results, with average values of 27.9453 dB in PSNR, 0.8048 in SSIM, and 0.8361 in FSIM. At the threshold level of T = 12, SCSOWOA obtained the highest performance, with SSIM and FSIM scores of 0.9340 and 0.9542, respectively. Furthermore, it demonstrated the lowest average execution time of 1.3221 s, offering up to a 40% improvement in computational efficiency compared with other metaheuristic methods. Conclusions: The SCSOWOA algorithm effectively balances exploration and exploitation processes, providing high-accuracy, low-variance, and computationally efficient segmentation. These findings highlight its potential as a robust and practical solution for AI-assisted histopathological image analysis and lung cancer diagnosis systems. Full article

The flourishing of the Internet of Things (IoT) has not only improved our lives in smart homes and healthcare but also made us more susceptible to cyberattacks. Legacy intrusion detection systems are simply overwhelmed by the scale and diversity of IoT traffic, which is why there is a need for more intelligent forensic solutions. In this paper, we present a statistical technique, the Averaging Detection Method (ADM), for detecting attack traffic. Furthermore, the five deep learning models SimpleRNN, LSTM, GRU, BLSTM, and BGRU are compared for malicious traffic detection in IoT network forensics. A smart home dataset with a simulated DoS attack was used for performance analysis of accuracy, precision, recall, F1-score, and training time. The results indicate that all models achieve high accuracy, above 97%. BiGRU achieves the best performance, 99% accuracy, precision, recall, and F1-score, at the cost of high training time. GRU achieves perfect precision and recall (100%) with faster training, which can be considered for resource-constrained scenarios. SimpleRNN trains faster with comparable accuracy, while LSTMs and their bidirectional counterparts are better at capturing long-term dependencies but are computationally more expensive. In summary, deep learning, especially BiGRU and GRU, holds great promise for boosting IoT forensic investigation by enabling real-time DoS detection and reliable evidence collection. Meanwhile, the proposed ADM is simpler and more efficient at classifying DoS traffic than deep learning models. Full article

Background/Objectives: Overweight and obesity is a continuing health concern for preadolescent youth. We assessed associations between sleep timing and sleep duration and body mass index/body composition in Latine youth. Methods: Participants were 119 Latine youth (mean age 11.53 year; 58.8% girls) and their mothers living in the rural Midwestern U.S. Youth reported their average bedtime and waking time. Heights and weights for children and mothers were measured by trained research assistants and were used to calculate BMI scores (in mothers), as well as BMI percentiles and overweight status (in youth). Mothers completed surveys for demographic variables. Results: Youth who went to bed before 9:30 PM (mean bedtime) obtained more sleep than those with later bedtimes (9.73 h vs. 8.63 h, respectively, t(117) = 7.88, p < 0.001). Each extra hour of sleep duration was associated with a decreased risk of being overweight (OR = 0.53 for weeknight sleep, OR = 0.67 for weekend night sleep), and each hour later to bed was related to increased risk for being overweight (OR = 2.35 on weeknights, and OR = 1.66 on weekend nights). To replicate previous work, we broke the youth up into four sleep timing groups: early-to-bed and early-to-rise (EE), early-to-bed and late-to-rise (EL), late-to-bed and early-to-rise (LE), and late-to-bed and late-to-rise (LL). Youth with LL sleep patterns on weeknights were much more likely to be overweight compared to youth with EE patterns (OR = 4.94). On weekend nights, compared to EE weekend youth, LE and LL weekend youth were more likely to be overweight (OR = 3.45 and OR = 3.32, respectively). Wake times were not significantly related to overweight risk. Conclusions: Sleep timing patterns, especially sleep duration and earlier bedtimes, may be important to address in future research on obesity interventions. Findings suggest that earlier bedtimes may play an important and complimentary role in health, in addition to sleep duration alone, and this study highlights the need for more research in underserved, minoritized populations. Full article

The problem of shock wave/boundary layer interaction induced by compression corners widely exists in the external and internal flows of various supersonic/hypersonic aircraft. In practical engineering applications, multistage continuous compression is often used in the fin/rudder structure, while in internal flow, multistage compression schemes are usually employed at the inlet to enhance total pressure recovery; therefore, it is necessary to investigate the characteristics of multistage compression corner shockwave/boundary layer interactions. In basic research, it is usually simplified as the double compression corner shockwave/boundary layer interaction issue. In this paper, an experimental study of hypersonic shock/boundary layer interaction characteristics is conducted under quiet and noise inflow conditions, respectively, for the double compression corner model. Using high-speed Schlieren, the typical structure of shockwave/shockwave interaction and shockwave/boundary layer interaction above the corner is explored under both quiet and noisy incoming flow conditions. Then, based on gray average, root-mean-square analysis, Fast Fourier transform, proper orthogonal decomposition, and dynamic mode decomposition methods, the time-average and unsteady characteristics of the double compression corner configuration-induced separation were studied, and a comparative analysis was conducted. The difference law between wind tunnel noise level and interaction characteristics was summarized. Finally, the characteristic length and spectral characteristics of unstable waves that dominated the stability of the plate boundary layer were studied. The formation mechanism of separation is discussed, which provides technical support for the internal and external aerodynamic design and targeted optimization of hypersonic vehicles. Full article

Forestry environments—such as logging sites, transport trails, and resource monitoring areas—are characterized by rugged terrain and irregularly distributed obstacles, which pose substantial challenges for AGV route planning. This poses challenges for route planning in automated guided vehicles (AGVs) and forestry machinery. To address these challenges, this study proposes a hybrid path optimization method that integrates an improved A* algorithm with the Dynamic Window Approach (DWA). At the global planning level, the improved A* incorporates a dynamically weighted heuristic function, a steering-penalty term, and Floyd-based path smoothing to enhance path feasibility and continuity. In terms of local planning, the improved DWA algorithm employs adaptive weight adjustment, risk-perception factors, a sub-goal guidance mechanism, and a non-uniform and adaptive sampling strategy, thereby strengthening obstacle avoidance in dynamic environments. Simulation experiments on two-dimensional grid maps demonstrate that this method reduces path lengths by an average of 6.82%, 8.13%, and 21.78% for 20 × 20, 30 × 30, and 100 × 100 maps, respectively; planning time was reduced by an average of 21.02%, 16.65%, and 9.33%; total steering angle was reduced by an average of 100°, 487.5°, and 587.5°. These results indicate that the proposed hybrid algorithm offers practical technical guidance for intelligent forestry operations in complex natural environments, including timber harvesting, biomass transportation, and precision stand management. Full article

Complications are frequent in cancer patients and contribute to adverse outcomes and higher healthcare costs, underscoring the need for earlier identification and prediction. This study evaluated the feasibility of using passively generated smartphone sensor data to explore early-warning signals of complications and symptom worsening during cancer treatment. A total of 108 patients were continuously monitored using accelerometer, GPS, and screen on/off data collected through the LAIMA application, while symptoms of depression, fatigue, and nausea were assessed every two weeks and complications were confirmed during clinic visits or emergency presentations. Smartphone data streams were aggregated into variables describing activity and sociability patterns. Machine learning models, including Decision Tree, Extreme Gradient Boosting, K-Nearest Neighbors, and Support Vector Machine, were used for complication prediction, and time-series models such as Autoregressive Integrated Moving Average, Holt–Winters, TBATS, Long Short-Term Memory neural network, and General Regression Neural Network were applied to identify early behavioral changes preceding symptom reports. In this exploratory analysis, the ensemble model demonstrated high sensitivity (89%) for identifying complication events. Smartphone-derived behavioral indicators enabled earlier detection of depression, fatigue, and vomiting by about nine days in a subset of patients. These findings demonstrate the feasibility of passive smartphone sensor data as exploratory early-warning signals, warranting validation in larger cohorts. Full article

Persistent patrolling with multiple Unmanned Aerial Vehicles (UAVs) remains challenging due to dynamic surveillance priorities, heterogeneous node importance, and evolving operational constraints. We present the novel Chimpanzee Troop Algorithm for Patrolling (CTAP), a decentralized policy inspired by chimpanzees fission–fusion dynamics and territorial behavior. CTAP provides three capabilities: (i) on-the-fly patrol-group instantiation, (ii) importance-aware territorial partitioning of the patrol graph, and (iii) adaptive boundary expansion via a lightweight shared-memory overlay that coordinates neighboring groups without centralization. Unlike the Ant Colony Optimization (ACO), Heuristic Pathfinder Conscientious Cognitive (HPCC), Recurrent LSTM Path-Maker (RLPM), State-Exchange Bayesian Strategy (SEBS), and Dynamic Task Assignment via Auctions (DTAP) baselines, CTAP couples local-idleness reduction with controlled edge-exploration, yielding stable coverage under shifting demand. We evaluate these approaches across multiple maps and fleet sizes using the average weighted idleness, global worst-weighted idleness, and Time-Normalized Idleness metrics. CTAP reduces the average weighted idleness by 7% to 22% and the global worst-weighted idleness by 30–65% relative to the strongest competitor and attains the lowest Time-Normalized Idleness in every configuration. These results show that a simple, communication-limited, partition-based policy enables robust, scalable patrolling suitable for resource-constrained UAV teams in smart-city environments. Full article

Background/Objectives: This study aimed to develop botanical nanofibers loaded with Psidium guajava leaf extract to heal wounds effectively. Methods: A 2³ factorial design was conducted to study the impact of freeze-drying parameters—freezing time, vacuum, and lyophilization time—on the total phenolic and flavonoid content in the lyophilized extract. Then, a polyurethane-based nanofiber dressing loaded with Psidium guajava leaf extract was fabricated using a one-step electrospinning technique. The nanofiber was evaluated considering total polyphenol and flavonoid content, surface roughness, and morphological assessment by scanning electron microscopy. Finally, the nanofiber was evaluated using in vivo wound-healing studies, histopathological analyses, and assessments of tissue levels of tumor necrosis factor-alpha, interleukin-6, matrix metalloproteinase, and growth factors. Results: The optimal conditions for freeze-drying the aqueous extract of Psidium guajava leaves were a freezing time of 24 h, a vacuum adjusted to 0.02 bar, and a lyophilization time of 48 h. The total polyphenol and flavonoid content within the nanofiber was 96 ± 1.2% and 91.83 ± 2.4%, respectively. Incorporating lyophilized extract in the nanofiber led to a decreased roughness average and root mean square roughness of the nanofiber. The nanofiber was continuous and had a smooth, uniform surface. The in vivo wound-healing assay showed superior wound-healing compared to the commercial Panthenol cream. These results were confirmed with histopathological studies. Conclusions: The extraction technique and lyophilization parameters significantly affect the bioactive content of Psidium guajava leaf extract. The botanical-loaded nanofiber showed greater wound-healing potential than a commercial cream, confirming its potential in regenerative medicine and wound repair applications. Full article

With the rapid growth of computation-intensive applications, high-performance computing (HPC) clusters have become essential for scientific computing, AI training, and industrial simulation. However, job scheduling in HPC clusters remains challenging due to heterogeneous resources, diverse task demands, and complex constraints. Traditional scheduling methods such as FCFS, SJF, and Backfilling show limited adaptability and struggle to achieve global optimization in large-scale environments. To address these issues, this paper proposes an intelligent scheduling method based on graph neural networks (GNNs) and deep reinforcement learning. A resource-constrained job–node bipartite graph is constructed to model task–node matching relationships, with node and task features capturing resource states and task demands. A GNN is employed to encode the scheduling state, and an Actor–Critic reinforcement learning framework is used to guide scheduling decisions. Simulation results show that, compared with other schedulers, the proposed GNN–Actor–Critic approach significantly improves average waiting time, average turnaround time, average slowdown, and overall resource utilization, demonstrating its effectiveness and practicality for HPC cluster scheduling. Full article

Vertical axial-flow pumps with bidirectional passages are widely used in applications requiring flow reversal. However, their unique inlet geometry often leads to asymmetric impeller inflow conditions. This study investigates the internal flow behavior and pressure pulsation characteristics of a vertical bidirectional axial-flow pump under design, critical stall, and deep stall conditions using unsteady Reynolds-averaged Navier–Stokes simulations combined with Fast Fourier Transform and wavelet analysis. Results show that the pump reaches peak efficiency at the design point, with critical and deep stall occurring at 0.6 Q_des and 0.5 Q_des, respectively. The head at the deep stall condition shows a further drop of 7.51% compared to the critical stall condition. This progressive performance degradation is attributed to vortex-induced blockage: it initiates with the intensification of the tip leakage vortex and evolves into large-scale separation vortices covering the suction surface under deep stall—a mechanism distinctly influenced by the bidirectional inlet’s stagnant water zone. Inlet asymmetry, reflected by a normalized velocity coefficient (V_n) below 0.6 in the stagnant water zone under design flow, is partially mitigated during stall due to flow confinement. Pressure pulsations at the blade leading edge are dominated by the blade passing frequency (BPF), with amplitudes under critical stall about 3.2 times those at design conditions. At the impeller outlet, critical stall produces a mixed dominant frequency (shaft frequency and BPF), whereas deep stall yields the highest pulsation amplitude (BPF ≈ 4.8 × the design value) resulting from extreme passage blockage. These findings clarify how bidirectional-inlet-induced vortices modulate stall progression and provide theoretical guidance for enhancing the operational stability of such pumps under off-design conditions. Full article

In froth flotation, the features of froth images are important information to predict the concentrate grade. However, the froth structure is influenced by multiple factors, such as air flowrate, slurry level, ore properties, reagents, etc., which leads to highly complex and dynamic changes in the image features. Additionally, issues such as the immeasurability of ore properties and measurement errors pose significant uncertainties including aleatoric uncertainty (intrinsic variability from ore fluctuations and sensor noise) and epistemic uncertainty (incomplete feature representation and local data heterogeneity) and generalization challenges for prediction models. This paper proposes an uncertainty quantification regression framework based on cross-modal interaction fusion, which integrates the complementary advantages of Selective Kernel Networks (SKNet) and Vision Transformers (ViT). By designing a cross-modal interaction module, the method achieves deep fusion of local and global features, reducing epistemic uncertainty caused by incomplete feature expression in single-models. Meanwhile, by combining adaptive calibrated quantile regression—using exponential moving average (EMA) to track real-time coverage and adjust parameters dynamically—the prediction interval coverage is optimized, addressing the inability of static quantile regression to adapt to aleatoric uncertainty. And through the localized conformal prediction module, sensitivity to local data distributions is enhanced, avoiding the limitation of global conformal methods in ignoring local heterogeneity. Experimental results demonstrate that this method significantly improves the robustness of uncertainty estimation while maintaining high prediction accuracy, providing strong support for intelligent optimization and decision-making in industrial flotation processes. Full article

On-demand public transport systems are increasingly adopted to improve service flexibility, reduce operating costs, and meet emerging mobility needs. Evaluating their performance under realistic demand and operational conditions, however, remains a complex challenge. This study presents a hybrid simulation framework that integrates deep learning-based demand forecasting, behavioural survey data, and agent-based simulation to assess system performance. A BiLSTM neural network trained on real-world smartcard data forecasts short-term passenger demand, which is embedded into an agent-based model simulating vehicle dispatch, routing, and passenger interactions. The framework is applied to a case study in Melbourne, Australia, comparing a baseline fixed-route service with two on-demand scenarios. Results show that the most flexible scenario reduces the average passenger trip time by 32%, decreases the average wait time by 34%, increases vehicle occupancy from 12.1 to 18.6 passengers per vehicle, lowers emissions per passenger trip by 72%, and cuts the service cost per trip from AUD 6.82 to AUD 4.73. These findings demonstrate the potential of hybrid on-demand services to improve operational efficiency, passenger experience, and environmental outcomes. The study presents a novel, integrated methodology for scenario-based evaluation of on-demand public transportation using real-world transportation data. Full article