Search Results (5,263)

Wheat grain protein content (GPC), quantified as the mass proportion of protein relative to the total dry grain weight, is a key indicator for wheat quality evaluation. Based on space–air–ground collaborative observations, this study screened GPC-sensitive spectral parameters and multi-stage and multi-scale remote-sensing GPC estimation models. Results showed that the optimal feature set derived from unmanned aerial vehicle (UAV) imagery consisted of four vegetation indices (VIs) and three optimized texture indices (TIs). The Random Forest-based Soil–Plant Analysis Development (SPAD) model for field plots achieved the highest accuracy at the anthesis stage, with a coefficient of determination (R²) of 0.91 and a root mean square error (RMSE) of 1.77. Cross-year validation gave a correlation coefficient of 0.80 and RMSE of 3.83. Data analysis indicated that GPC of mature wheat had an extremely significant correlation with SPAD values across various growth stages, higher than leaf area index (LAI). We thus established a quantitative GPC estimation framework with UAV-derived SPAD as an intermediate variable, which kept stable R² (0.60) and RMSE (1.1%) except during grain-filling. On this basis, UAV-retrieved GPC data were scaled up to 10 m. Eight optimal parameters were selected from Sentinel-2 satellite data, and the regional-scale GPC estimation model was developed via the partial least squares regression (PLSR) algorithm. Compared with the model established directly using ground measured data without scale conversion, the scale-up model increased R² by 8.9% and reduced RMSE by 40%, improving the model inversion accuracy. The proposed regional-scale GPC remote-sensing estimation model effectively solves the scale mismatch between ground observation and satellite data, and provides technical support for field and regional wheat quality remote-sensing estimation. Full article

With the development of Internet of Things- and unmanned aerial vehicle (UAV)-based power inspection, the accurate and efficient detection of insulator defects has become the key to the safe and stable operation of transmission lines. However, in real UAV inspection scenarios, insulator defect detection is still faces many challenges, such as complex backgrounds, large-scale variations, small defect regions, and weak fault-related features. Existing lightweight detection models are often difficult to achieve a good balance between detection accuracy and computational efficiency. To address these problems, we propose an insulator defect detection method, YOLOv12n-HEPPSA, which combines the Pooling Partial Self-Attention (PPSA) mechanism with evolutionary neural architecture search (ENAS). A PPSA module is introduced at the end of the backbone network of YOLOv12n to enhance its ability to represent high-level features, so as to capture the differences between insulator defect regions and their adjacent normal structures. On this basis, an evolutionary search is performed by optimizing detection accuracy and computational cost. In addition, a historical-memory-guided evolutionary search and evaluation strategy is designed to reduce the randomness of the search process and improve optimization efficiency of candidate architectures. The experimental results obtained using the publicly available Unified Insulator Public Dataset (UPID) and four external datasets demonstrate that the proposed method balances detection accuracy and computational efficiency while exhibiting strong lightweight features and transferability across diverse UAV-based power inspection scenarios. Full article

The rapid decarbonisation of urban districts is accelerating the deployment of rooftop photovoltaic (PV) systems, household battery energy storage systems (BESSs) and electric vehicles (EVs). However, the high variability of urban micro-generation creates substantial forecasting errors at the urban–grid interface, imposing high balancing costs on distribution system operators (DSOs) and local energy communities. Traditional metrics fail to assess how effectively internal peer-to-peer (P2P) flexibility offsets these forecast mismatches prior to grid settlement. To address this gap, this paper presents a novel methodological framework and a non-parametric indicator: the Community Imbalance Neutralisation Index (CINI). Formulated in a generalised, scalable matrix structure applicable to any heterogeneous urban neighbourhood (N households), CINI quantifies the relative reduction in net community-level imbalance relative to the cumulative sum of uncoordinated individual forecast errors. CINI ranges from 0 per cent (no collective mitigation) to 100 per cent (perfect internal neutralisation). Complementing this index, an adaptive day-ahead scheduling algorithm is introduced to determine the optimal community energy purchase requirement (E_forecast). The proposed framework is numerically evaluated using a high-resolution synthetic benchmark annual dataset with 15 min intervals (35,040 intervals) representing a Central European urban residential cluster equipped with diverse combinations of PV, BESS and managed EV charging infrastructure. The simulation results demonstrate that active cVPP coordination reduces annual grid-facing imbalance energy from 188.73 MWh to 143.88 MWh, increasing the annual CINI score from 57.22% to 67.39% (+10.17 percentage points) compared with the uncoordinated baseline. Notably, the framework reveals a ‘Winter Flexibility Paradox’, achieving its highest relative efficacy during the winter months (+13.88 percentage points in December). Furthermore, sensitivity analyses show that scaling flexibility up to 40 kW achieves a CINI score of 91.22%, revealing diminishing marginal returns and critical technological saturation thresholds. The proposed CINI metric and E_forecast dispatch algorithm provide city planners, municipal energy managers and DSOs with a transparent diagnostic tool to design dynamic socio-economic tariff incentives, optimise urban micro-grid sizing, prevent free-rider dynamics, and foster resilient, self-balancing smart cities. Full article

This study introduces a formative assessment approach for examining changes in cognitive demand across student responses using open-ended questions and rubric-based evaluation. The method is designed for small cohorts and emphasizes instructional refinement across predefined cognitive levels. It integrates Generative AI to support rubric application, scoring consistency, and identification of response patterns without replacing instructor judgment or established assessment practices. The approach addresses challenges in specialized programs with low enrollment, where large-scale statistical designs may be impractical, particularly in rapidly evolving fields such as automotive engineering. The methodology was applied in a vehicle dynamics course with 19 students, self-assigned to 6 teams of similar size by selecting an available slot; teams were then allocated to one of two learning paths by team number, in which the same activities were delivered in reversed order in a crossover design combining simulation-based and experiential activities. Assessments were aligned with Bloom’s taxonomy and ABET outcomes. Four Generative AI models (GPT-4 Turbo, GPT o4-mini, DeepSeek, and GPT-4o-mini) were used to support scoring under rubric and prompting procedures defined by an experienced instructor. Overall, inter-model agreement was moderate (Krippendorff’s $\alpha =0.666$). This value did not meet the reference criterion of 0.7 adopted in the study and indicates meaningful model-dependent variation in rubric application. Within this limitation, task-level performance indicators showed a shift toward higher cognitive demand, particularly at the application level, with responses moving from conceptual definition toward early analysis. The approach also surfaced learning gaps that informed instructional adjustments, supporting its potential as a diagnostic tool for continuous improvement in engineering education. Full article

Image disturbances caused by platform ego-motion are coupled with true target motion under the dynamic viewpoints of unmanned ground vehicles (UGVs), resulting in trajectory-prediction drift, association mismatches, and persistent contamination of identity information by unreliable observations. This paper proposes ERM-Track, a disturbance-aware and reliability-guided online multi-object tracking framework. I${}^2$DF-Mamba uses a causal dual-stream Mamba encoder to integrate IMU, joint-state, and trajectory histories and predicts a trajectory-specific image-disturbance distribution. The predicted disturbance mean and uncertainty are mapped explicitly from normalized/log-scale disturbance coordinates to the detection-observation space. CF-TUR then estimates observation reliability through reliable–contaminated posterior fusion and causal evidence accumulation and generates separate bounded write gains for the motion state and identity memory. On the 6488-frame sealed holdout set of the additionally annotated CEAR data, ERM-Track obtains 64.34% HOTA, 66.28% AssA, and 73.42% IDF1, with 28 identity switches. Three-seed backbone replacement experiments show that Mamba provides the highest mean HOTA among the evaluated causal encoders. Post hoc isotonic calibration reduces ECE from 0.4752 to 0.0478, with only marginal changes in the tracking metrics. The complete pipeline reaches 69.50 FPS on an RTX 4080 workstation and an onboard mean latency of approximately 29 ms on a Jetson Orin NX, corresponding to a reciprocal processing rate of approximately 34.5 FPS. Full article

Bio-inspired integrity monitors for automated vehicles are reported as single pooled detection figures, which conflates algorithmic performance with evaluation protocol. Transferring the reafference principle to inertial-channel integrity, we identify what actually governs the reported figure. A single-track forward model is identified from the data, and its residual is standardized by Operating-Condition Residual Normalization (OCRN), a bin-wise operator conditioned on an observable operating point; the design is resolved by an exhaustive constrained search and each component is isolated using ablation. Across 224,638 samples spanning five towns and four friction levels, the residual scale varies by a factor of 158, and OCRN recovers 21.8 F₁ points, more than the detection rule, the encoder, and the biological attenuation gate combined. A cross-comparison confirms this: changing the detection rule moves the result by 0.09 points, while changing the normalization moves this by 15 to 19. The monitor attains 97.34% precision, 76.35% recall, and 85.57% F₁ at a 0.76% false-alarm rate when counting every injection, and 97.30/93.99/95.62% above a 3σ detectability floor when covering 80.1% of them. The floor scales with forward-model error, which is correlated with but not determined by the road friction, and the gate, the most explicitly biological element, is not selected once the residual is conditionally normalized, with its best setting gaining at most one F₁ point at nearly twice the false-alarm rate. Full article

Person-overboard detection from maritime unmanned aerial vehicle (UAV) imagery is challenging because the targets occupy very few pixels, sea-surface clutter is severe, human appearance varies substantially, and real-time processing is required. Existing detectors therefore struggle to meet the demands of practical maritime search and rescue. This paper presents HPSNet, an accuracy- and recall-oriented detector designed for maritime UAV imagery. HPSNet uses YOLOv11 as its baseline and introduces three complementary modifications. A channel transposed attention (CTA) module is embedded in the backbone to improve the discrimination of target features from complex sea-surface interference. A Giraffe feature pyramid network (GFPN) replaces the original feature-fusion network to strengthen multiscale information exchange and preserve cues from extremely small targets. A Dynamic Head (DyHead) adapts the predictions to variations in target scale, location, and appearance. HPSNet is evaluated against 12 representative detectors on the public Person Detection in Water and AFO datasets, and ablation experiments examine the contribution of each component. On Person Detection in Water, HPSNet achieves 78.2% mAP@50, 37.5% mAP@50:95, and 67.2% recall, improving the YOLOv11 baseline by 2.7, 2.2, and 4.2 percentage points, respectively. On AFO, it achieves 88.4% mAP@50 and 59.4% mAP@50:95, with gains of 0.7 and 1.9 percentage points over the baseline. The model contains 4.18 M parameters, requires 9.6 GFLOPs, and processes an image in 13.4 ms on an RTX 4090. These results demonstrate improved detection of small and visually weak maritime targets relative to YOLOv11 while maintaining a moderate model scale, providing a foundation for future deployment and optimization on embedded maritime UAV platforms. Full article

Unmanned aerial vehicle (UAV) operational training is often constrained by high equipment costs, safety risks, and limited training venues. In addition, novice operators may experience stress and anxiety in safety-risk environments, adversely affecting attention, decision-making, and task performance. To overcome these challenges, this study developed a virtual UAV training system using Unity and C#. The system enables bidirectional command communication and real-time state synchronization between virtual and physical UAVs through UDP-based communication. A quasi-experimental design was employed to evaluate the system’s technical feasibility and training performance. Sixty university students without prior UAV experience were assigned to either the virtual or physical training group. Learning performance was assessed in terms of knowledge acquisition, flight performance, learning motivation, cognitive load, and technology acceptance. Results showed that both groups demonstrated significant improvements in knowledge and flight performance, with statistical/practical equivalence within a prespecified margin after training, indicating comparable learning achievement. The virtual training group obtained higher “Relevance” scores on the ARCS motivation scale but experienced greater extraneous cognitive load, whereas the physical training group showed higher technology acceptance. Overall, the proposed system provides a safe, flexible, and effective alternative to conventional physical UAV training. Future work may optimize the user interface to reduce extraneous cognitive load and enhance technology acceptance, improving its applicability in skills-oriented education and training. Full article

In SAE Level 3 automated driving, drivers are required to supervise vehicle operation and respond to overtaking requests. Owing to task monotony and insufficient workload, drivers are prone to passive fatigue, which may impair vigilance and safety. This study investigated passive fatigue development during automated driving and proposed a multimodal detection method. Thirty licensed participants completed one automated driving task and one manual driving task in a driving simulator. Eye-movement and ECG/heart rate variability indicators were synchronously collected, and fatigue states were assessed using the Karolinska Sleepiness Scale. Results show that passive fatigue during automated driving developed differently from active fatigue during manual driving. Based on PERCLOS, pupil diameter, pupil diameter variation, SDNN, and LF/HF, an SVM-based passive fatigue detection model was developed to classify alert and passive fatigue states. Across repeated subject-wise validation, the model achieved an accuracy of 89.19%, sensitivity of 91.83%, specificity of 86.54%, balanced accuracy of 89.19%, F1 score of 89.48%, and precision of 87.28%, outperforming the model trained on manual driving active fatigue data. These findings demonstrate the need for scenario-specific driver-state monitoring models in automated driving systems and provide an applied physiological sensing approach for passive fatigue detection and warning design. Full article

City-scale microscopic traffic simulation is increasingly used for policy evaluation, operational planning, and disruption analysis, where repeated scenario runs must retain vehicle-level trajectories rather than only aggregate traffic indicators. This requirement creates an efficiency–consistency trade-off: single-node simulators preserve centralized state ownership but become inefficient for million-vehicle tasks, whereas distributed execution reduces runtime but may disrupt vehicle updates at partition boundaries. Cross-partition movement can cause trajectory breaks, duplicate or missing updates, and inconsistent local interaction states if boundary events, vehicle context, and update ownership are not coordinated. To address this problem, this study proposes DiCoSim, a distributed coordination framework for boundary-consistent large-scale microscopic traffic simulation. DiCoSim integrates incremental spectral-clustering partitioning, spatio-temporal event aggregation, and acknowledgment-controlled state handoff to coordinate workload balance, boundary communication, and vehicle handoff. Experiments on a 483 km${}^2$ Tianjin network with 1.5 million agents show that DiCoSim achieved a $14.49\times$ strong-scaling speedup on 16 compute nodes while maintaining close agreement with centralized execution. For the fixed boundary-crossing evaluation cohort, the trajectory interruption rate was reduced to 0.06%. In addition, a single 72 h continuous high-load run achieved 99.96% availability. These results indicate that, under the tested Tianjin conditions, coordinated boundary management supports efficient million-agent microscopic simulation while maintaining vehicle-state continuity across partitions. Full article

Unmanned aerial vehicle (UAV) acquired imagery provides a flexible non-contact visual sensing modality for monitoring inland waterway locks, yet reliable perception remains challenging due to significant scale variation among targets, as well as partial visibility, low contrast, water-surface texture variation, and complex backgrounds. To address these issues, this study proposes BISRF-Net, a baseline-preserving scale-guided residual feature routing network built upon the CBv2 Faster R-CNN framework. The method retains the original backbone, feature pyramid network, and region proposal network, while introducing scale-guided residual routing at the region-of-interest (ROI) refinement stage. Through identity-preserved residual addition, the baseline ROI representation is preserved and enhanced with complementary scale-sensitive information without altering the original feature pathway. Experiments conducted on the UAV image subset of the TROUT lock-monitoring dataset demonstrate that BISRF-Net maintains comparable overall detection performance while improving medium-scale target representation. Repeated trials with different random seeds further assess the robustness of the proposed method and indicate that its main benefit lies in medium-scale target refinement. Ablation and computational analyses further show that the proposed design enables targeted ROI-level refinement with additional computational cost. These findings highlight the potential of scale-guided ROI feature refinement for robust UAV-based visual sensing in complex inland waterway environments. Full article

Large-scale public events generate concentrated passenger demand during egress periods, often overwhelming urban transit systems. This paper proposes a multi-modal evacuation framework that coordinates in-service buses temporarily diverted from existing lines and dedicated shuttle vehicles pre-positioned at depots. The problem is formulated as a two-layer stochastic optimization under travel time uncertainty: the upper layer determines pre-event shuttle fleet sizing, while the lower layer makes real-time dispatching decisions for both modes. We propose an Uncertainty-Aware Reinforcement Learning framework with Categorical DQN (UARL-CD) that learns a robust dispatching policy through a reward function aligned with the lower-level objective, explicitly accounting for travel time uncertainty via distributional value representation and stochastic training, with an action masking mechanism enforcing operational constraints. Simulation experiments based on a realistic stadium evacuation scenario demonstrate that the proposed framework significantly outperforms deterministic optimization and rule-based strategies, achieving a 31.6% reduction in evacuation completion time and a 48.4% reduction in average passenger waiting time compared to shuttles alone, while maintaining robustness to travel time uncertainty with only 4.0% performance degradation and online decisions executed within the 2-min decision interval. Full article

Objects in images captured by unmanned aerial vehicles (UAVs) are often small in scale, weak in texture, and frequently occluded. Meanwhile, complex backgrounds can produce local responses similar to those of real targets, increasing the risk of missed detections for small objects and false detections in background regions. To address these challenges, this paper proposes DPFS-YOLO, a small-object detection method designed for complex aerial scenes. Built upon YOLOv8n, the proposed method integrates detail enhancement, foreground selection, and semantic guidance into a collaborative optimization framework, aiming to improve small-object feature representation while suppressing spurious background responses. First, the Dual-Path Edge Fusion (DPEF) module combines explicit edge priors, Gaussian smoothing constraints, and a learnable detail modeling branch to enhance object boundaries and local texture representations. Second, the Foreground-Guided Detail Selection (FGDS) module filters the enhanced detail responses through foreground gating, preserving target-relevant information while weakening false activations caused by complex backgrounds. Finally, the Semantic-Aware Guided Fusion for Tiny Object Detection (SAG-tiny) branch adds a P2 high-resolution detection layer and fuses shallow spatial details with top–down high-level semantic information before semantic-guided refinement, enabling small-object details to be constrained by semantic context during detection. Experiments on the VisDrone2019-DET, HIT-UAV, and TinyPerson datasets demonstrate that the proposed method effectively alleviates missed detections of small objects and reduces false-detection risk under complex background conditions. Full article

Boaters use class ratings to describe whitewater difficulty on a scale from I to VI, with ratings assigned and refined through the judgment of experienced boaters. These practices are effective, but the addition of physical measurements could improve their reliability and facilitate whitewater comparisons across different rivers and flows. This exploratory study considers specific stream power (SSP), a physics-based metric describing energy transfer in rivers, as an indicator of whitewater difficulty. Data were collected at eight study sites on the Upper Deschutes River in Oregon using a camera-mounted DJI Phantom 4 RTK quadcopter. Sites were assessed at one or two flows and whitewater difficulty ranged from Class I flatwater to Class V cascading rapids. Stream slope and width data were derived from a combination of unmanned aerial vehicle (UAV) photogrammetry and aerial light detection and ranging (LiDAR), and SSP was calculated for each site. Whitewater class ratings were strongly associated (df = 6, p < 0.05) with site average SSP, r_s= 0.95, 95% CI [0.72, 1.00], site average slope, r_s = 0.95, 95% CI [0.72, 1.00], and within-site maximum SSP, r_s = 0.93, 95% CI [0.58, 1.00]. The within-site maximum slope, minimum width, and constriction ratio had significant but smaller relationships. Physical assessments of whitewater conditions may help support management decisions in dam removal, hydropower relicensing, or other instream flow negotiations. Full article

Predicting driver-state evolution in extra-long tunnel corridors remains challenging because of prolonged spatial confinement and repeated lighting transitions. This study uses a statistical human–vehicle composite, the comprehensive driving index (CDI), as a reproducible quantitative target for predictive auditing. In this study, multimodal information denotes synchronized ocular, physiological, vehicle-motion, and environmental sensor signals; the objective is to quantify their incremental predictive value rather than introduce a new fusion architecture. Fully nested leave-one-driver-out cross-validation with a prespecified 120 s unsupervised initialization estimated all preprocessing, scaling, PCA, model-selection, and calibration parameters from training data only. The five components explained 60.36% of target variance. In the original-range 30 s task (4835 evaluation windows), history-only ridge regression achieved an RMSE of 0.08294 and an R² of 0.166, while directly tuned AR achieved an RMSE of 0.08261 and an R² of 0.168. On the common 4259-window sample, expanded ridge and AR achieved RMSEs of 0.08123 (R² 0.187) and 0.08153 (R² 0.182). Adding coarse scene information produced an ΔRMSE = +0.00002 (95% CI −0.00026 to 0.00027), whereas external environmental summaries produced an ΔRMSE = −0.00019 (95% CI −0.00037 to −0.00003). HistGradientBoosting did not improve performance. The primary contribution is a leakage-controlled predictive-audit framework for screening candidate information sources before deployment decisions. Full article