Search Results (396)

Remaining Useful Life (RUL) prediction for aeroengines represents a core challenge in Prognostics and Health Management (PHM), with significant implications for condition-based maintenance, operational cost reduction, and flight safety enhancement. Current deep learning-based approaches encounter three major limitations when handling multi-source sensor data: conventional convolution operations struggle to model heterogeneous sensor feature distributions, leading to computational redundancy; simplistic multimodal fusion strategies often induce semantic conflicts; and high model complexity hinders industrial deployment. To address these issues, this paper proposes a novel Partial Convolution Attention-enhanced CNN-LSTM Network (PALC-Net). We introduce a partial convolution mechanism that applies convolution to only half of the input channels while preserving identity mappings for the remainder. This design retains representational power while substantially lowering computational overhead. A dual-branch feature extraction architecture is developed: the temporal branch employs a PConv-CNN-LSTM architecture to capture spatio-temporal dependencies, while the statistical branch utilizes multi-scale sliding windows to extract physical degradation indicators—such as mean, standard deviation, and trend. Additionally, an adaptive fusion module based on cross-attention is designed, where heterogeneous features are projected into a unified semantic space via Query-Key-Value mappings. A sigmoid gating mechanism is incorporated to enable dynamic weight allocation, effectively mitigating inter-modal conflicts. Extensive experiments on the NASA C-MAPSS dataset demonstrate that PALC-Net achieves state-of-the-art performance across all four subsets. Notably, on the FD003 subset, it attains an MAE of 7.70 and an R² of 0.9147, significantly outperforming existing baselines. Ablation studies validate the effectiveness and synergistic contributions of the partial convolution, attention mechanism, and multimodal fusion modules. This work offers an accurate and efficient solution for aeroengine RUL prediction, achieving an effective balance between engineering practicality and algorithmic sophistication. Full article

This study develops a machine-learning-based framework for multiclass soil classification using Cone Penetration Test (CPT) data, aiming to overcome the limitations of traditional empirical Soil Behavior Type (SBT) charts and improve the automation, continuity, robustness, and reliability of stratigraphic interpretation. A dataset of 340 CPT soundings from 26 sites in Shanghai is compiled, and a sliding-window feature engineering strategy is introduced to transform point measurements into local pattern descriptors. An XGBoost-based multiclass classifier is then constructed using fifteen engineered features, integrating second-order optimization, regularized tree structures, and probability-based decision functions. Results demonstrate that the proposed method achieves strong classification performance across nine soil categories, with an overall classification accuracy of approximately 92.6%, an average F1-score exceeding 0.905, and a mean Average Precision (mAP) of 0.954. The confusion matrix, P–R curves, and prediction probabilities show that soil types with distinctive CPT signatures are classified with near-perfect confidence, whereas transitional clay–silt facies exhibit moderate but geologically consistent misclassification. To evaluate depth-wise prediction reliability, an Accuracy Coverage Rate (ACR) metric is proposed. Analysis of all CPTs reveals a mean ACR of 0.924, and the ACR follows a Weibull distribution. Feature importance analysis indicates that depth-dependent variables and smoothed p_s statistics are the dominant predictors governing soil behavior differentiation. The proposed XGBoost-based framework effectively captures nonlinear CPT–soil relationships, offering a practical and interpretable tool for high-resolution soil classification in subsurface investigations. Full article

Vehicle tracking is essential for autonomous driving, traffic surveillance, and intelligent transportation, yet most existing trackers rely on frame-level training that neglects temporal dependencies. This mismatch between training and testing leads to error propagation, mislocalization in challenging frames, and failure to re-identify vehicles after occlusion. We present a reinforcement learning (RL)-based sequence-level training framework that formulates tracking as a sequential decision process and directly incorporates evaluation metrics consistent with testing. Our approach enhances robustness in difficult frames and occlusion scenarios by leveraging temporal decision dependencies, and introduces a temporal data augmentation strategy based on sliding-window sampling to improve generalization across diverse motion patterns. Experiments on challenging benchmarks indicate that our method provides improved robustness and temporal continuity over frame-level training approaches, suggesting the benefits of incorporating sequence-level learning in vehicle tracking. Full article

The emerging development of Global Navigation Satellite System (GNSS) software receivers has opened new opportunities in diverse operations. However, non-line-of-sight (NLOS) concatenated signal reception is one prevalent deterioration factor causing positioning errors in urban scenarios. To enhance integrity and reliability through receiver autonomous integrity monitoring (RAIM) techniques in urban environments, distinguishing between line-of-sight (LOS) and NLOS signals facilitates the exclusion of NLOS channels: this is challenging due to uncertain signal reflections/refractions from diverse obstruction conditions in the built environment. Moreover, NLOS features show similarity to multipath effects like scattering and diffraction which causes difficulty in identifying the NLOS type. Recent work exploited NLOS detections with multi-correlator outputs using neural networks that outperform using signal strength techniques for NLOS detection. This paper proposes a neural network approach designed to recognise and learn spatial features among early, late, and prompt correlator outputs, differentiating between correlations, and also by memorising temporal features to acquire propagation information. Specifically, the spatial features of correlator IQ streams are derived from convolutional layers incorporated with concatenations, to formulate associate models like early-minus-late discrimination. A Recurrent Neural Network (RNN), i.e., long short-term memory (LSTM), is integrated to obtain comprehensive temporal features; hereby, a softmax classifier is appended in the last layer to distinguish between NLOS and LOS signals. By simulating synthetic datasets generated by a Spirent simulator and captured by a software-defined radio (SDR), the correlator outputs are acquired during the scalar tracking stage. The product of the proposed network demonstrates high performance in terms of accuracy, time consumption and sensitivity, affirming the efficiency of utilising early-stage correlations for NLOS detection. Moreover, an impact analysis of varying the sliding window length on NLOS discrimination underscores the need to fine-tune the parameter, as well as balancing accuracy, operation complexity and sensitivity. Full article

In the Arctic region, the navigation and positioning accuracy of shipborne and autonomous underwater vehicle (AUV) integrated Global Navigation Satellite System (GNSS) and Inertial Navigation System (INS) solutions is severely degraded due to poor satellite geometry, frequent ionospheric disturbances, non-Gaussian measurement noise, and strong multipath effects, as well as long-term INS-based dead-reckoning for AUVs when GNSS is unavailable underwater. In addition, the sparse ground-based augmentation infrastructure and the lack of reliable reference trajectories and dedicated test ranges in polar waters hinder the validation and performance assessment of existing marine navigation systems, further complicating the achievement of accurate and reliable navigation in this region. To improve the positioning accuracy of the GNSS/INS shipborne navigation system, this paper adopts a tightly coupled GNSS/INS navigation approach. To further enhance the accuracy and robustness of tightly coupled GNSS/INS positioning, this paper proposes an improved Adaptive Robust Extended Kalman Filter (IAREKF) algorithm to effectively suppress the effects of gross errors and non-Gaussian noise, thereby significantly enhancing the system’s robustness and positioning accuracy. First, the residuals and Mahalanobis distance are calculated using the Adaptive Robust Extended Kalman Filter (AREKF), and the chi-square test is used to assess the anomalies of the observations. Subsequently, the observation noise covariance matrix is dynamically adjusted to improve the filter’s anti-interference capability in the complex Arctic environment. However, the state estimation accuracy of AREKF is still affected by GNSS signal degradation, leading to a decrease in navigation and positioning accuracy. To further improve the robustness and positioning accuracy of the filter, this paper introduces a sliding window mechanism, which dynamically adjusts the observation noise covariance matrix using historical residual information, thereby effectively improving the system’s stability in harsh environments. Field experiments conducted on an Arctic survey vessel demonstrate that the proposed improved adaptive robust extended Kalman filter significantly enhances the robustness and accuracy of Arctic integrated navigation. In the Arctic voyages at latitudes 80.3° and 85.7°, compared to the Loosely coupled EKF, the proposed method reduced the horizontal root mean square error by 61.78% and 21.7%, respectively. Full article

In response to the challenges posed by the complex evolution of risks and the static nature of traditional assessment methods during the construction of long-span prestressed concrete (PC) continuous rigid-frame bridges, this study proposes a risk assessment framework that integrates visual perception with dynamic probabilistic reasoning. By combining an improved YOLOv8 model with the Decision-making Trial and Evaluation Laboratory–InterpretiveStructure Modeling (DEMATEL–ISM) algorithm, the framework achieves intelligent identification of risk elements and causal structure modelling. On this basis, a dynamic Bayesian network (DBN) is constructed, incorporating a sliding window and forgetting factor mechanism to enable adaptive updating of conditional probability tables. Using the Tongshun River Bridge as a case study, at the identification layer, we refine onsite targets into 14 risk elements (F1–F14). For visualization, these are aggregated into four categories—“Bridge, Person, Machine, Environment”—to enhance readability. In the methodology layer, leveraging causal a priori information provided by DEMATEL–ISM, risk elements are mapped to scenario probabilities, enabling scenario-level risk assessment and grading. This establishes a traceable closed-loop process from “elements” to “scenarios.” The results demonstrate that the proposed approach effectively identifies key risk chains within the “human–machine–environment–bridge” system, revealing phase-specific peaks in human-related risks and cumulative increases in structural and environmental risks. The particle filter and Monte Carlo prediction outputs generate short-term risk evolution curves with confidence intervals, facilitating the quantitative classification of risk levels. Overall, this vision-guided dynamic risk assessment method significantly enhances the real-time responsiveness, interpretability, and foresight of bridge construction safety management and provides a promising pathway for proactive risk control in complex engineering environments. Full article

This study addresses the problem of correcting systematic errors in classical cascade object detectors under severe data scarcity and distribution shift. We focus on the widely used Viola–Jones framework enhanced with a modified Census transform and propose a modular “corrector” architecture that can be attached to an existing detector without retraining it. The key idea is to exploit the blessing of dimensionality: high-dimensional feature vectors constructed from multiple cascade stages are transformed by PCA and whitening into a space where simple linear Fisher discriminants can reliably separate rare error patterns from normal operation using only a few labeled examples. This study presents a novel algorithm designed to correct the outputs of object detectors constructed using the Viola–Jones framework enhanced with a modified census transform. The proposed method introduces several improvements addressing error correction and robustness in data-limited conditions. The approach involves image partitioning through a sliding window of fixed aspect ratio and a modified census transform in which pixel intensity is compared to the mean value within a rectangular neighborhood. Training samples for false negative and false positive correctors are selected using dual Intersection-over-Union (IoU) thresholds and probabilistic sampling of true positive and true negative fragments. Corrector models are trained based on the principles of high-dimensional separability within the paradigm of one- and few-shot learning, utilizing features derived from cascade stages of the detector. Decision boundaries are optimized using Fisher’s rule, with adaptive thresholding to guarantee zero false acceptance. Experimental results indicate that the proposed correction scheme enhances object detection accuracy by effectively compensating for classifier errors, particularly under conditions of scarce training data. On two railway image datasets with only about one thousand images each, the proposed correctors increase Precision from 0.36 to 0.65 on identifier detection while maintaining high Recall (0.98 → 0.94), and improve digit detection Recall from 0.94 to 0.98 with negligible loss in Precision (0.92 → 0.91). These results demonstrate that even under scarce training data, high-dimensional feature separation enables effective one-/few-shot error correction for cascade detectors with minimal computational overhead. Full article

With the continuous development of cities and the increasing utilization of underground space, ambient noise seismic imaging has become an essential approach for exploring and monitoring the urban subsurface. The integration of Distributed Acoustic Sensing (DAS) with ambient noise imaging offers a more convenient and effective solution for investigating shallow subsurface structures in urban environments. To overcome the limitations of conventional ambient noise seismic nodes, which are costly and incapable of achieving high-density data acquisition, this work makes use of existing urban telecommunication fibers to record ambient noise and perform sliding-window cross-correlation on it. Then the Phase-Weighted Stack (PWS) technique is applied to enhance the quality and stability of the cross-correlation signals, and fundamental-mode Rayleigh wave dispersion curves are extracted from the cross-correlation functions through the High-Resolution Linear Radon Transform (HRLRT). In the inversion stage, an adaptive regularization strategy based on automatic L-curve corner detection is introduced, which, in combination with the Preconditioned Steepest Descent (PSD) method, enables efficient and automated dispersion inversion, resulting in a well-resolved near-surface S-wave velocity structure. The results indicate that the proposed workflow can extract useful surface-wave dispersion information under typical urban noise conditions, achieving a feasible level of subsurface velocity imaging and providing a practical technical means for urban underground space exploration and utilization. Full article

Periapical radiographs remain the first-line imaging modality in endodontics due to accessibility and low radiation dose, whereas cone-beam computed tomography (CBCT) is reserved for inconclusive cases or suspected anatomical complexity. We propose a physics- and geometry-aware preprocessing pipeline coupled with sliding-window Viterbi tracking to enhance canal visibility and recover plausible root canal trajectories directly from routine periapical images. The pipeline standardizes row-wise brightness, compensates for the cone-like tooth density profile (Tukey window), and suppresses noise prior to dynamic-programming inference, requiring only minimal operator input (two-point orientation and region of interest). In a retrospective evaluation against micro-computed tomography (micro-CT)/CBCT reference anatomy, the approach accurately localized canals on periapicals under study conditions, suggesting potential as a rapid, chairside aid when 3D imaging is unavailable or deferred. Full article

Water quality monitoring is crucial for public health and environmental protection, but traditional methods lack real-time accuracy. This study addresses this gap by combining feature augmentation methods (e.g., sliding window) with Principal Component Analysis (PCA) and a tailored Two-Layer Regularized Gated Recurrent Unit (TLR-GRU) model for efficient water quality prediction. Results demonstrate that the proposed framework significantly improves prediction accuracy, with R² increased by 7.78%, RMSE decreased by 27.36%, MAE by 36.71%, and MAPE by 45.08%. This approach offers a novel technical pathway for real-time environmental monitoring and water resource management. Full article

Smart cities increasingly rely on intelligent technologies to improve urban infrastructure, sustainability, and quality of life. Traffic flow prediction is essential for the optimization of the transportation system, reducing congestion and improving mobility. However, real-world traffic data are often noisy, limited in size, and lack sufficient features to capture the flow dynamics and temporal dependencies, making accurate prediction a significant challenge. Previous studies have shown that recurrent neural network (RNN) variants, such as LSTM and GRU, are well-suited for time series forecasting tasks, but their performance is highly sensitive to hyperparameter settings. This study proposes a hybrid approach that integrates GRU with a metaheuristic optimization algorithm to address this challenge. After effective preprocessing steps and a sliding time window are applied to structure the data, particle swarm optimization (PSO) is utilized to optimize the hyperparameters of the GRU. The model’s performance is evaluated using RMSE, MAE, and $R^2$, and compared against several baseline approaches, including LSTM, CNN-LSTM, and a manually configured GRU. According to the experimental findings, the GRU model that was manually adjusted performed the best overall. However, the PSO-GRU model demonstrated competitive results, confirming that metaheuristics offer a promising alternative when manual tuning is not feasible despite the higher computational costs. Full article

Ischemic heart disease (IHD) remains the predominant cause of global mortality, necessitating accurate incidence forecasting for effective prevention strategies. Existing statistical models inadequately capture nonlinear epidemiological patterns, while deep learning approaches lack clinical interpretability. We constructed an interpretable predictive framework combining particle swarm optimization (PSO), bidirectional long short-term memory (BiLSTM) networks, and a novel multi-scale attention mechanism. Age-standardized incidence rates (ASIRs) from the Global Burden of Disease (GBD) 2021 database (1990–2021) were stratified across 24 sex-age subgroups and processed through 10-year sliding windows with advanced feature engineering. SHapley Additive exPlanations (SHAP) provided a three-level interpretability analysis (global, local, and component). The framework achieved superior performance metrics: mean absolute error (MAE) of 0.0164, root mean squared error (RMSE) of 0.0206, and R² of 0.97, demonstrating a 93.96% MAE reduction compared to ARIMA models and a 75.99% improvement over CNN–BiLSTM architectures. SHAP analysis identified females aged 60–64 years and males aged 85–89 years as primary predictive contributors. Architectural analysis revealed the residual connection captured 71.0% of the predictive contribution (main trends), while the BiLSTM–Attention pathway captured 29.0% (complex nonlinear patterns). This interpretable framework transforms opaque algorithms into transparent systems, providing precise epidemiological evidence for public health policy, resource allocation, and targeted intervention strategies for high-risk populations. Full article

The accurate and rapid estimation of ionospheric delays is essential for PPP-RTK positioning. While traditional spatial interpolation methods like Kriging rely solely on geographic correlations, they often fail to capture rapid temporal variations in the ionosphere. To overcome this limitation, this paper proposes a long short-term memory (LSTM)-based interpolation method for interpolating ionospheric delays between satellites. The method leverages both spatial and short-term temporal correlations to generate accurate ionospheric corrections at user locations. The model uses a sliding window approach, taking the most recent 10 min of historical data as input to predict ionospheric delays at the current epoch. Experimental validation using data from a reference network in Australia—with average and maximum baseline lengths of 280 km and 650 km, respectively—demonstrates that the proposed LSTM method achieves a centimeter-level interpolation accuracy, with RMS errors between 0.06 m and 0.07 m under both quiet and geomagnetic storm conditions, significantly outperforming the Kriging method (0.27–0.44 m). In PPP-RTK, the LSTM model achieved a 3D positioning accuracy of 8.99 cm RMS during quiet periods, representing improvements of 51.9% and 28.8% over the No Constraint and Kriging methods, respectively. Under geomagnetic storm conditions, it maintained a 3D RMS of 24.54 cm—over 44% more accurate than other methods—and reduced the average time-to-first-fix (TTFF) to just 7.0 min, a 39.1% improvement. This study provides a novel approach for ionospheric spatial interpolation, demonstrating a particular robustness even during geomagnetic storms. Full article

Autonomous exploration is essential for various mapping tasks, including data collection, environmental monitoring, and search and rescue operations. Unmanned aerial vehicles (UAVs), owing to their low cost and high maneuverability, have become key enablers of such applications, particularly in complex or hazardous environments. However, existing approaches often suffer from issues such as redundant exploration and unstable flight behavior. In this study, we propose a hierarchical exploration approach specifically designed for limited-field-of-view UAVs in geospatial mapping applications. The approach addresses these challenges through hybrid viewpoint generation, an innovative boundary exploration sequence, and a two-stage global path planning strategy. To balance exploration efficiency and computational cost, we adopt a hybrid approach that combines collision-free spherical sampling with adaptive viewpoint generation based on stochastic differential equations. This approach generates high-quality candidate viewpoints while minimizing computational overhead. Furthermore, we introduce a novel heuristic evaluation function to prioritize frontiers within small regions, thereby facilitating optimal path planning. Based on this formulation, the global coverage path is modeled as a traveling salesman problem (TSP). The two-stage global planning framework consists of an initial stage that applies a history-aware trajectory enhancement strategy with smoothing corrections, followed by a second stage employing a sliding-window TSP algorithm to construct the global path. This design mitigates motion inconsistencies caused by frequent heuristic updates and enhances flight stability and trajectory smoothness. To evaluate the performance of the proposed framework, we compare it with state-of-the-art approaches in both simulated and real-world environments. Experimental results demonstrate that our approach shortens flight paths and reduces exploration time, thereby improving overall exploration efficiency. Full article

Aggregated sales forecasting is an important research topic in the highly competitive retail industry. With the availability of different data sources, various techniques and models have been proposed for aggregated sales forecasting. However, existing methods often overlook the spatial correlation in sales between neighboring retailers. In this paper, we propose a new framework for aggregated sales forecasting based on the deep learning technique ConvLSTM with an attention mechanism to solve this challenge. In the new framework, ConvLSTM is utilized to fully leverage spatially relevant information from adjacent retailers, while the attention mechanism is employed to capture spatial dependencies and select the most pertinent data from spatial inputs. Furthermore, a spatial sliding window technique is designed to augment the sample size. To validate the efficacy of our proposed framework, we conducted experiments using real-world retail sales data and compared our model with established benchmarks. Additionally, we conducted an ablation study to assess the contributions of key components, including the attention mechanism and spatial data augmentation. The experimental results demonstrate that our proposed model effectively improves the prediction performance, offering a novel approach to aggregate sales forecasting for both industry and academia. Full article