Search Results (95)

Against the backdrop of the advancing Transportation Power Strategy, long and large tunnels face critical challenges in ensuring the safety and efficiency of transportation scheduling due to their harsh environment, complex traffic network, and the need for coordination among multiple types of vehicles. Addressing the shortcomings of existing research—such as the disconnection between path planning and dynamic environments, insufficient coordination between timetables and paths, and incomplete conflict management—this paper constructs a comprehensive optimization model for the scheduling of construction vehicles in tunnel traffic networks. Firstly, integrating the improved social force model with the BPR function, an adaptive social force-BPR path planning model with a collision compensation mechanism is proposed, and the weights of sub-items are optimized using the improved AHP algorithm. Secondly, a constraint system covering paths, spatio-temporal logic, and three types of conflicts (crossing conflicts, head-on conflicts, and congestion conflicts) is established, and a bi-objective function of “minimum total scheduling time” and “minimum number of conflicts” is designed. Combined with the improved NSGA-II algorithm, the collaborative optimization of departure intervals and paths is realized. In particular, a conflict entropy repair operator is introduced to quantify the conflict chaos through node conflict entropy and vehicle conflict entropy, and the scheduling strategy is accurately adjusted based on the logic of “priority ranking-dynamic delay” to balance conflict resolution and efficiency loss. Finally, a case verification is carried out relying on a tunnel topological network with 30 nodes and 41 edges. The experimental results show that the optimal repulsion coefficient k_f of the social force model is 20, and the maximum departure interval of 8 min is the best configuration after introducing the repair operator. At this time, the total scheduling time is 136 min, and the total number of conflicts is only 2, completely avoiding high-risk head-on conflicts and congestion conflicts. The research outputs a vehicle scheduling scheme, enriches the theory of tunnel traffic scheduling, and provides scientific and feasible technical support for the coordinated scheduling of construction vehicles in long and large tunnels. Full article

This paper develops a unified asymptotic theory for conditional single-index U-statistics and the associated conditional U-processes in the setting of locally stationary functional time series subject to missing-at-random response mechanisms. The proposed framework addresses, within a single nonparametric inferential architecture, three major sources of complexity in modern functional data analysis: infinite-dimensional covariates, smoothly time-varying stochastic dynamics, and incomplete response observations. The methodology is based on a class of kernel-type estimators combining temporal localization, functional single-index smoothing, and inverse-propensity correction. Temporal localization captures the gradual evolution of the underlying regression structure, the single-index projection provides an effective dimension-reduction mechanism for functional covariates, and the propensity adjustment restores the target conditional functional under the MAR sampling scheme. The principal contribution of the paper is the establishment of weak convergence, in a suitable space of bounded functions, for the resulting propensity-adjusted conditional U-process indexed by a general class of measurable kernels. Under absolute regularity conditions, local stationarity assumptions, small-ball probability requirements, entropy restrictions of VC type, and uniform consistency of the propensity-score estimator, the normalized process is shown to converge weakly to a tight centered Gaussian process. The limiting covariance structure explicitly reflects the interaction between temporal smoothing, functional concentration, dependence, and the random loss of responses. In parallel, uniform convergence rates are derived for the associated conditional single-index U-statistic estimators, thereby quantifying the respective contributions of smoothing bias, stochastic fluctuation, local-stationarity approximation error, and missingness-induced variance inflation. A substantial part of the analysis is devoted to the technical difficulties created by the simultaneous presence of dependence, nonstationarity, functional covariates, and incomplete observations. The proofs combine Hoeffding-type decompositions adapted to weighted incomplete data, blocking and coupling arguments for absolutely regular triangular arrays, refined entropy bounds for kernel-indexed function classes, and small-ball probability techniques for functional covariates. The MAR mechanism is incorporated via inverse-propensity weighting, and its effects on the effective sample size, asymptotic variance, and bias structure are made explicit. The theory also provides a rigorous foundation for bandwidth selection through blocked, propensity-adjusted cross-validation and clarifies its relation to the corresponding oracle risk. The proposed framework encompasses a broad class of statistical learning and inference problems involving pairwise or higher-order functionals of functional time series. In particular, it applies to conditional Kendall-type functionals, discrimination problems, metric learning with incomplete labels, and conditional independence testing under local stationarity. A simulation study illustrates the finite-sample behavior of the proposed estimators and supports the theoretical findings across varying regimes of temporal nonstationarity, serial dependence, functional concentration, and response missingness. Overall, the results provide a mathematically rigorous and methodologically flexible foundation for inference from evolving functional data when dependence, infinite dimensionality, and incomplete observation are present simultaneously. Full article

Early fault prediction of wind turbines is critical for ensuring wind farm safety and reducing operation and maintenance costs. However, the latent and progressive nature of incipient faults, together with concurrent failures across multiple subsystems, makes accurate root-cause identification challenging. In addition, severe class imbalance between normal and faulty samples further degrades prediction performance, particularly for minority fault types. To address these challenges, this paper proposes a novel fault prediction model, M²ANet, using SCADA data within a 30-min pre-fault window. The model combines a dual-memory module with progressive dilated convolutions to efficiently capture multi-scale temporal dependencies from high-dimensional operational variables. An entropy-bias penalty is further introduced into the loss function to adaptively regularize the predicted probability distribution, alleviating overconfidence under imbalanced data conditions and improving the recognition of minority faults. Experiments on a real-world wind farm dataset show that M²ANet achieves an overall accuracy of 90.73% and a weighted F1-score of 90.62% in multi-class fault prediction, outperforming 10 representative baseline models. In addition to these aggregate metrics, per-class evaluation confirms the model’s robustness under class imbalance. Notably, for yaw system faults, which account for only 1.9% of the samples, M²ANet achieves a recall of 95.92% with a 30-min-ahead warning. These results demonstrate its effectiveness and reliability for early fault prediction in practical wind turbine applications. Full article

To improve the accuracy and stability of short-term traffic flow prediction in complex road networks and address the limitations of existing models in modeling spatiotemporal dependencies, this paper proposes a traffic flow prediction model based on a Hypergraph Spatio-Temporal Interaction Network (HGSTIN) in the context of intelligent transportation systems. The study constructs a multi-dimensional traffic pattern input tensor by integrating three temporal scales—proximity, intra-day, and intra-week—while taking traffic flow as the prediction target and introducing average speed and lane occupancy as auxiliary features. In terms of temporal modeling, a Transformer architecture integrated with a Dynamic Tanh (DyT) mechanism is adopted to capture multi-period variations. For spatial modeling, a neighborhood hypergraph and a DTW-based semantic hypergraph are combined to enhance the representation of local and global through spatial self-attention and hypergraph neural network branches, and an adaptive feature fusion module is designed to perform adaptively weighted fusion of the outputs from the two branches. In terms of loss function design, a temporal gradient consistency loss function is proposed to enhance the robustness of predictions. Experimental results on the PEMS04 and PEMS08 datasets show that the proposed model achieves average improvements of approximately 5.15%, 1.76%, and 3.88% in MAE, RMSE, and MAPE, respectively, compared to the second-best baseline model. The model exhibits the smallest performance degradation in multi-step prediction scenarios, and the effectiveness of each module is validated through ablation studies. The findings demonstrate that HGSTIN can effectively capture the dynamic spatiotemporal characteristics of complex traffic scenarios, thereby providing high-precision prediction support for intelligent transportation systems. Full article

Electrocardiogram (ECG) analysis is a cornerstone non-invasive diagnostic technique for detecting cardiac arrhythmias, which remain a leading cause of mortality worldwide. While recent advances in deep learning have significantly improved automated arrhythmia classification, the current literature lacks systematic, fair comparisons of fundamental neural architectures under unified experimental conditions, and very few studies provide model interpretability. This study addresses these gaps by first providing a rigorous comparative analysis of three representative architectures—Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and Residual Network (ResNet)—on the MIT-BIH Arrhythmia Database under identical preprocessing, training, and evaluation protocols. We then propose an efficient Fine-Tuned CNN (FT-CNN) optimized for ECG signal characteristics through adaptive kernel sizing for P-QRS-T morphological extraction, multi-faceted regularization including L2, dropout, and batch normalization, cosine annealing learning rate, and a custom loss function combining weighted categorical cross-entropy with focal loss with gamma equal to 2.0 to address severe class imbalance. The FT-CNN achieves an accuracy of 98.51%, outperforming fourteen benchmark models, including standard CNN with an accuracy of 97.20%, ResNet with 96.88%, LSTM with 96.50%, GRU with 96.30%, and traditional classifiers. Comprehensive ablation studies confirm an improvement of 6.17% over the baseline. Class-wise analysis reveals excellent performance for normal beats with an F1-score of 0.99, ventricular ectopic beats with 0.95, and unknown beats with 0.98, while supraventricular ectopic beats with an F1-score of 0.79 and fusion beats with 0.70 remain challenging. Unlike most prior works, we integrate Grad-CAM and Integrated Gradients for explainability, quantitatively evaluating attribution faithfulness, sanity checks, and noise robustness. Full article

As intelligent cockpits transition into the “third living space”, traditional driver monitoring systems face limitations such as rigid monitoring, computationally intensive algorithms, and insufficient engineering robustness. This paper proposes an edge-computing-based emotion-aware ambient lighting system, forming a complete loop of emotion perception–decision–adaptation. A lightweight emotion recognition network is designed for edge computing: the Mini_XCEPTION architecture is optimized with depthwise separable convolutions to reduce parameters, and a Gaussian-smoothed weighted cross-entropy loss function is used to address class imbalance and ambiguous emotion boundaries. After INT8 quantization, the model achieves 47 FPS real-time inference on a Raspberry Pi (Raspberry Pi Ltd., Cambridge, United Kingdom). A high-concurrency asynchronous software–hardware architecture based on PyQt5 5.15.6 and QThread5.15.6 is built, with a serial communication mechanism featuring fixed-length frames and fault recovery to improve the robustness of the hardware-in-the-loop system. Breaking the rigid alarm mode, an emotion–HSV lighting mapping matrix is established based on the Russell Valence-Arousal model, combined with 0.1 Hz bionic breathing rhythm for non-intrusive feedback. An FSM-controlled HSV lighting policy with 0.1 Hz breathing-light feedback was implemented on an in-cabin HIL platform. In a 12-participant simulated road-rage test, the intervention reduced FER-based anger recovery time by 42.6%; independent physiological validation remains necessary. Full article

The rapid integration of photovoltaic (PV) generation into active distribution networks (ADNs) creates a fundamental tension between maintaining tight voltage regulation and accommodating high distributed energy resource (DER) penetration levels. Conventional voltage control methods such as the droop control operate locally without coordination, while centralized optimal power flow requires full network observability and reliable real-time communication. Multi-agent deep reinforcement learning (MADRL) methods provide adaptive coordination but suffer from long training times and algorithmic complexity that prevent direct deployment on embedded inverter hardware. This paper proposes the Optimal Historical Selection and Forecasting (OHSF) scheme: a physics-informed long short-term memory (LSTM) network combined with an online sensitivity-based correction loop for medium-voltage ADNs. A composite loss function incorporating data-driven regression, an inter-PV voltage sensitivity penalty, and an inverter capability constraint produces reactive power setpoints that are inherently aware of physical limits, while the correction loop refines the predictions using real-time AC power flow feedback. The OHSF scheme supports a centralized full-network mode and a decentralized fallback mode in which the trained weights run locally on each inverter. Simulations under worst-case PV placement and network reconfiguration on the modified IEEE 33-bus and 69-bus test systems achieve an average voltage deviation across all PV buses of $0.701\%$ and $0.601\%$ at $172\%$ DER penetration on the 33-bus system, and $0.804\%$ and $0.806\%$ at $242\%$ DER penetration on the 69-bus system, while training 32 to 49 times faster than state-of-the-art MADRL methods. Full article

Graph Neural Networks (GNNs) have achieved remarkable progress in community detection, which is an essential topic in network analysis with the aim of dividing a network into multiple subgraphs to mine potential information. However, most existing GNN-based community detection approaches adopt static loss function weights during the training process. In this paper, we propose an unsupervised end-to-end community detection framework and define an adaptive loss weighting layer within this framework, named QALW, which is capable of learning an optimal combination of loss weights during model training to balance reconstruction loss and clustering loss. Experimental results obtained based on three real-world benchmark datasets (Cora, Citeseer, and Pubmed) demonstrate that QALW achieves effective and stable community detection performance compared with eight representative baseline methods. In particular, QALW improves ACC by 5.7% over the strongest baseline on the Cora dataset, and achieves ACC values of 64.9%, 61.7%, and 63.9% on Cora, Citeseer, and Pubmed, respectively. Furthermore, the results verify that the proposed dynamic scheduling mechanism effectively alleviates gradient conflicts and enables more stable optimization than fixed-weight strategies. Overall, QALW demonstrates promising competitiveness and good robustness for unsupervised community detection in attributed networks. Full article

Agriculture faces significant challenges from crop diseases, which threaten global food security and cause substantial economic losses annually. While deep learning has advanced plant disease detection, existing models often struggle with generalization across heterogeneous environments and real-time deployment constraints, hindering their practical application in diverse agricultural settings. This paper proposes AD-DETR, an enhanced real-time detection transformer framework specifically designed for agricultural scenarios. The model incorporates three key innovations to address these issues. First, the Multi-Scale Align Network (MSANet) achieves adaptive feature alignment through an Adapt Fusion Align (AFA) block, effectively preserving disease detail information across varying scales. Second, the Spatial–Spectral Attentive Feature Fusion (SSAFF) module integrates frequency-domain processing with attention mechanisms, enhancing feature representation quality by combining spatial and spectral information. Third, the IPIoUv2 loss function improves bounding-box regression accuracy through an internal perception mechanism and scale-adaptive weighting. Comprehensive experiments demonstrate that AD-DETR achieves strong performance, with 90.2% mean average precision at $IoU=0.5$ on the Crop Disease dataset and 97.4% on the PlantDoc dataset. It maintains high efficiency with 16.4 million parameters, 47.2 GFLOPs computational complexity, and inference speeds of 230–242 frames per second. These results indicate that AD-DETR is robust to domain shift and suitable for resource-constrained applications, such as real-time monitoring on mobile and edge platforms. Full article

This research examines how false data injection attacks (FDIAs) impact the trajectory tracking control of underactuated surface vessels (USVs). The internal uncertain dynamics of the system are reconstructed using radial basis function neural networks (RBFNNs). In order to avoid the computational pressure of the RBFNNs on the system, the neural network weights, external disturbances, and FDIAs are converted into a single parameter learning form using the minimum learning parameters (MLPs). Next, a nonlinear feedback function is constructed and introduced into the controller design process, thereby avoiding the controller accuracy loss caused by MLPs. Within the backstepping method framework, the adaptive laws leverage deep information robust adaptive technology to estimate the upper limits of the uncertainty term. The closed-loop system is provided with a rigorous theoretical analysis by combining the Lyapunov stability theory. Finally, the effectiveness of the control scheme is verified by simulation. The results show that the proposed controller guarantees boundedness of all closed-loop signals and drives the tracking errors into a small neighborhood of the reference trajectory even under the attack of FDIAs and the influence of internal and external uncertainties. Full article

Isothermal dispatch models for battery energy storage systems (BESSs) systematically underestimate degradation costs because dispatch-induced Joule heating elevates cell temperature and accelerates ageing through Arrhenius-type kinetics. This paper proposes three integrated contributions. First, a thermal–electrochemical coupling loop embeds a first-order lumped thermal model within the dispatch simulation: cell temperature is updated from I²R heat generation and Newton cooling at each time step, and the resulting temperature trajectory feeds into the Arrhenius stress factors of a semi-empirical degradation model combining $\sqrt{\Delta t}$-based calendar ageing with Rainflow-based cycle ageing, enabling the optimiser to discover thermally self-regulating strategies. This coupling is critical because, as the results demonstrate, ignoring it leads to systematic underestimation of degradation costs by up to 13%. Second, the resulting four-objective problem (negative profit, thermally coupled degradation cost, SOC deviation, and CVaR imbalance penalty) is solved by a hypervolume-contribution-enhanced NSGA-III (HVC-NSGA-III), which augments reference-point selection with an archive pruned by removing the solution of the smallest individual hypervolume contribution, concentrating Pareto resolution in the knee region. Third, an SOH-adaptive knee-point selection assigns the degradation weight as a monotone function of ageing degree $(1-\mathrm{SOH})/(1-{\mathrm{SOH}}_{\mathrm{EOL}})$, automatically tightening dispatch conservatism as remaining useful life diminishes. Simulations on ENTSO-E data over 96 h show the following: (i) thermal coupling shifts the Pareto front by 8–15% in the degradation dimension with temperature excursions up to 7 K; (ii) HVC-NSGA-III improves hypervolume by 8.7% over standard NSGA-III; (iii) SOH-adaptive selection reduces capacity loss by 27.4% at only 9.1% revenue cost; and (iv) ablation confirms Rainflow (24.8%) and thermal coupling (13.1%) as the two largest contributors. Full article

In complex real-world scenarios, object detection faces significant challenges due to severe noise interference and feature degradation. To overcome these limitations, this paper proposes DFA-YOLO, an enhanced YOLOv11 framework integrating three key innovations. First, a Deformable Spatial Attention (DSA) module is introduced into the C3k2 backbone blocks, which dynamically adjusts the receptive field to focus on informative spatial regions. This significantly enhances the model’s adaptability to geometric variations and occluded objects. Second, a Hierarchical Multi-Scale Fusion Module (HMFM) is designed to dynamically recalibrate feature responses across scales, enhancing the model’s perception of multi-scale targets. Third, an improved Wasserstein loss function combines small-object adaptive weighting with dynamic gradient modulation to address boundary ambiguity and scale sensitivity under adverse conditions. Extensive experiments on the RTTS dataset validate the superiority of our approach, achieving improvements of 3.4% and 2.8% in mAP50 and mAP50-95, respectively. Additional experiments on the Exdark dataset confirm the method’s robust generalization capability, with significant accuracy gains observed across all benchmarks. Full article

In disease control and precision management in aquaculture, rapid and accurate identification of common fish diseases is pivotal to mitigating economic losses and ensuring aquaculture profitability. However, fish diseases are characterized by subtle symptoms, polymorphic lesions, and high susceptibility to environmental perturbations such as water turbidity and illumination fluctuations. Existing detection models generally suffer from inadequate lightweight design, poor fine-grained lesion feature extraction, and deficient adaptability to class imbalance, failing to meet the stringent requirements of precise diagnosis in real-world aquaculture scenarios. To address these challenges, this study proposes FDR-Net: a fine-grained lesion detection model for tilapia via multi-scale feature enhancement and spatial attention fusion. Using image data of Nile tilapia (Oreochromis niloticus) covering 6 common diseases and healthy individuals (from the NTD-1 dataset), the model incorporates symmetry-aware design logic, leveraging the morphological and textural symmetry of healthy tilapia tissues to capture lesion-induced symmetry-breaking features, thereby improving fine-grained lesion detection accuracy. Through depth-width scaling coefficients, FDR-Net achieves lightweight optimization while integrating three core modules and a task-specific loss function for full-chain optimization: specifically, a Micro-lesion Feature Enhancement Module (MLFEM) is embedded in key feature layers of the backbone network to accurately extract edge and texture features of incipient fine-grained lesions via multi-scale frequency decomposition and residual fusion; subsequently, a Lightweight Multi-scale Position Attention Module (MS_PSA) and a Single-modal Intra-feature Contrastive Fusion Module (SMICFM) are collaboratively deployed—the former focusing on spatial localization of lesion features, and the latter enhancing lesion-background discriminability through channel-spatial feature recalibration and contrastive fusion; finally, a Class-Aware Weighted Hybrid Loss (CAWHL) function is combined with customized small-target anchor boxes to alleviate class imbalance and further improve localization and classification accuracy of fine-grained lesions. Empirical evaluations on the NTD-1 dataset demonstrate that compared with mainstream state-of-the-art baseline models, FDR-Net achieves a peak recognition accuracy of 90.1% with substantially enhanced mAP_50-95 performance. Retaining lightweight characteristics, it exhibits superior performance in identifying incipient fine-grained lesions and strong adaptability to simulated complex aquaculture scenarios. Collectively, this study provides an efficient technical backbone for the rapid and precise detection of tilapia fine-grained lesions, offering a potential solution for precise disease management in tilapia farming. Full article

Numerous studies have demonstrated the potential of Synthetic Aperture Radar (SAR) in monitoring glacier velocity. However, owing to the complex dynamics of glaciers and the variability of their surface features, velocity fields derived from even short-interval SAR image pairs often exhibit missing parts. This study proposes a missing glacier velocity recovery method based on a spatiotemporal hybrid neural network to solve the above problem. Considering the spatiotemporal characteristics of glacier velocity fields, a hybrid network combining an Artificial Neural Network (ANN) and a Denoising Autoencoder (DAE) is developed. The ANN is first employed to capture spatial correlations associated with missing values, after which it is integrated with the DAE to model temporal variations using a time-aware loss function. An iterative weighting strategy adaptively balances spatial and temporal features during training. The method is applied to SAR–derived velocity fields of Petermann Glacier. Experimental results show that the method significantly improves the performance of glacier velocity recovery compared to traditional methods. Additionally, the study compares and analyzes the velocity of Petermann Glacier in different seasons, and the findings indicate that the glacier exhibits more pronounced seasonal differences in the accumulation zone. Full article

Accurate urban green space (UGS) mapping is essential for assessing urban ecosystem health and supporting sustainable development planning. However, deep learning-based UGS segmentation from Red–Green–Blue (RGB) remote sensing imagery faces two major challenges. First, the absence of near-infrared (NIR) information in RGB imagery hinders the ability to discriminate spectrally similar classes, such as vegetation and non-vegetation. Second, conventional convolutions with fixed receptive fields struggle to model the complex and irregular boundaries characteristic of UGS. To address these challenges, this study combined the Normalized Green–Red Difference Index with the Deformable Convolutional Network Lab (NGRDI-DCNLab) model, a semantic segmentation model tailored specifically for RGB-only imagery. Based on the DeepLabV3+ framework, the model introduced three core improvements: (1) The Normalized Green–Red Difference Index (NGRDI) was incorporated to compensate for the absence of NIR information, enhancing the spectral separability of vegetation pixels. (2) Standard convolutions in the decoder were replaced with deformable convolutions, enabling the network to more effectively adapt to irregular boundaries of UGS. (3) An NGRDI-weighted loss function was designed to assign higher weights to challenging samples and uncertain boundary regions, guiding the model toward more accurate edge delineation. Comprehensive evaluations on two public high-resolution datasets—the Wuhan Dense Labeling Dataset (WHDLD) and the Beijing subset of the Urban Green Space-1m dataset (UGS-1m_Beijing)—demonstrated that the NGRDI-DCNLab model outperformed existing popular deep learning models (like Unet++, etc.). Specifically, the deformable convolution effectively enhances the feature modeling capability for irregular boundaries; incorporating the NGRDI vegetation index as a fourth channel strengthens spectral feature representation and improves the distinction between vegetation and non-vegetation; and adding the dynamic NGRDI-weighted loss enables targeted learning for challenging samples. Through the synergistic effect of these three modules, the model achieves mean Intersection over Union (MIoU) scores of 84.77% and 77.66%, as well as F1-scores of 91.75% and 87.27%, on the WHDLD and UGS-1m_Beijing datasets, respectively. Furthermore, the model exhibited certain generalization capability on the unmanned aerial vehicle (UAV) dataset, the Urban Drone Dataset 6 (UDD6), attaining an MIoU of 87.43%. Our results confirm that high-precision UGS extraction is achievable using only RGB remote sensing imagery, providing a cost-effective and practical technical solution for refined urban governance and ecological monitoring. Full article