Search Results (789)

To address the issue of persistent high-magnitude outlier interference affecting long-baseline (LBL) positioning systems in complex marine environments, this paper proposes a kinematic constraint-based Robust Interacting Multiple Model Kalman Filter algorithm. Combined with anchor point initialization and multi-step historical observations, the algorithm constructs a spatial Euclidean distance discriminant criterion. By further incorporating the maximum velocity constraint of the Autonomous Underwater Vehicle (AUV), dynamic decision thresholds are established, and final detection decisions are output to the positioning system. Within the Kalman Filter recursion process, a measurement mask matrix is introduced to instantly isolate measurement outliers, preventing abnormal data from participating in state updates and model probability evolution. Simulation results demonstrate that, compared with standard LBL positioning, conventional single outlier detection, and the conventional maximum correntropy criterion-based Kalman filter (MCC-KF) algorithm, the proposed approach enhances outlier identification and suppression—particularly under consecutive anomaly conditions—thereby improving the positioning accuracy of maneuvering targets in complex underwater scenarios. Full article

In the LoRa (long-range) wide area network (LoRaWAN), Class A devices employ a pure ALOHA random access mechanism. Under large-scale access and bursty traffic conditions, severe packet collisions are likely, which reduces throughput and increases the packet loss rate. To address these issues, herein, we propose a collision mitigation scheme integrating the extended Kalman filter (EKF) with nonorthogonal multiple access (NOMA). First, a nonlinear state-space model is constructed to capture the dynamic evolution of backlog nodes and the uncertainty of traffic arrivals. The backlog node number is modeled as the hidden state, while newly arrived and successfully decoded packets are incorporated into the state-transition equation. At the gateway, decoded packet counts and channel occupancy are treated as observations based on which a nonlinear mapping between system state and observable features is established. The EKF is then applied to recursively predict and correct, enabling real-time estimation of the backlog state. Accordingly, an adaptive backoff strategy is designed to adjust transmission probability based on the estimated optimal load. Furthermore, to mitigate packet loss caused by collisions, a power-domain NOMA scheme with successive interference cancelation (SIC) is introduced. Signals transmitted with different spreading factors (SFs) are decoupled into approximately independent processing branches by exploiting inter-SF quasi-orthogonality. To account for imperfect inter-SF orthogonality, cross-SF residual coupling coefficients are introduced to characterize leakage interference. For transmissions sharing the same SF, overlapping packets are successively decoded and recovered through a NOMA-SIC mechanism jointly constrained by the SINR-based decoding threshold, the power-domain separation requirement, the maximum number of resolvable SIC layers, and residual SIC interference. Accordingly, the proposed receiver architecture enhances the decoding and recovery capability for collided LoRa packets. Simulation results demonstrate that, under medium-to-high traffic loads, the proposed scheme significantly improves throughput and access success rate while effectively reducing collision probability and packet loss, thereby enhancing the overall robustness and efficiency of the LoRa network. Full article

Reporting delays are a common challenge in actuarial reserving and infectious disease surveillance, where incomplete development information can distort real-time estimation and decision-making. Classical reserving methods, such as the chain ladder method, assume stable development patterns across event periods, which may be unrealistic when reporting behavior evolves over time. This paper develops a Hoerl-type state-space framework, in which development dynamics evolve as latent stochastic processes within a linear Gaussian state-space model. Estimation is conducted using the Kalman filter and Rauch-Tung-Striebel smoother, allowing recursive estimation under incomplete run-off triangles. The paper further establishes consistency and asymptotic normality for estimators of latent states, ultimate quantities, and the effective reproduction number. Simulation and empirical applications show that the proposed method performs comparably to Mack’s model under stable development patterns while providing substantially more accurate estimates of effective reproduction numbers when reporting behavior varies over time or delays remain unresolved near the boundary of the observation window. These results suggest that the proposed approach provides a flexible and theoretically grounded extension of classical actuarial reserving methods. Full article

Dynamic industrial processes often exhibit complex variable interactions, and time-varying behaviors, which pose significant challenges to conventional multivariate statistical monitoring methods. To address these issues, this paper proposes a novel data-driven monitoring framework that integrates knowledge-informed bipartite graph embedding with multi-scale dissimilarity analysis. First, a bipartite graph-embedding strategy is developed to incorporate mechanistic knowledge into the modeling process, enabling a more interpretable representation of dynamic relationships among process variables. On this basis, a multi-scale recursive dissimilarity monitoring method is further designed to enhance detection performance by capturing process variations across different temporal scales while reducing sensitivity to sliding window selection. The effectiveness of the proposed framework is validated through a numerical example and a benchmark simulation process. The results demonstrate that the proposed method achieves improved fault detection performance and robustness compared with conventional approaches. Full article

Thickness accuracy is a critical quality indicator in medium and heavy plate production, as it directly affects material utilization, product performance, and manufacturing cost. The rolling process of medium and heavy plates is highly nonlinear. It also involves multivariable coupling and dynamic fluctuations in operating conditions. Therefore, achieving highly accurate and reliable thickness prediction in industrial applications remains a major challenge. To address this issue, this paper develops a joint point-interval prediction framework for medium and heavy plate thickness in industrial applications. First, recursive feature elimination with a LinearSVR estimator (LinearSVR-RFE) is employed to eliminate low-contribution features from the original process feature set, retain informative variables, and construct a compact and effective feature subset. Second, Bayesian optimization is employed to tune the hyperparameters of multiple machine learning regression models. A Stacking ensemble strategy is then adopted to improve the accuracy and robustness of point prediction under complex production conditions. Finally, quantile regression is introduced based on the optimal point prediction model to construct prediction intervals at multiple confidence levels. This provides uncertainty-aware results for production decision-making. Experimental results based on real industrial data from a 3500 mm medium and heavy plate production line show that the proposed framework achieves strong point prediction performance on the test set. The optimal Stacking model achieves a coefficient of determination (R²) of 0.9845 with a root mean square error (RMSE) of 0.73 mm on the test set. In addition, the framework produces prediction intervals with a good balance between coverage and compactness at confidence levels from 80% to 95%. For example, at the 90% confidence level, the interval prediction module achieves a PICP of 0.9043 and a PINAW of 0.0711. The results indicate that the proposed framework provides an effective solution for intelligent thickness prediction and quality evaluation in industrial rolling processes. It also shows good potential for engineering applications. Full article

This paper presents the mechatronic design, mathematical modeling, parameter identification, and nonlinear position control of an open-architecture biaxial shake table capable of generating base acceleration along two orthogonal horizontal directions. The shake table is tailored for engineering research and education. Addressing the limitations of proprietary “black-box” systems, the platform is constructed using standard industrial components (HLTNC-CNC modules and NEMA 23 BLDC motors) to ensure reproducibility. A core contribution is the characterization of the system’s nonlinear dynamics to enhance tracking fidelity. The mathematical model, derived via the Euler–Lagrange formulation, incorporates viscous and Coulomb friction phenomena, which are critical for accurately reproducing zero-velocity crossings in seismic signals. System parameters are identified using the Recursive Least Squares (RLS) algorithm combined with State Variable Filters (SVFs) to process the regression vector. To enable precise closed-loop performance, a nonlinear state observer incorporating the identified friction dynamics is designed for velocity estimation. Furthermore, a Computed Torque Control (CTC) strategy is synthesized and compared against a conventional Proportional-Velocity (PV) controller. Experimental validations using historical ground motions, including the 1986 Colima earthquake, confirm that the CTC strategy reduces the maximum absolute tracking error by more than 75% compared to the PV approach, bounding the peak error to $0.36\mathrm{mm}$ across both axes. Furthermore, in high-amplitude scenarios, the proposed model-based approach achieved an RMS tracking error reduction of more than 83%. These results validate the proposed platform as a reliable and accessible tool for structural dynamics testing. Full article

Short food supply chains (SFSCs) have gained attention as mechanisms for strengthening local food systems and enhancing producer value. However, many SFSCs involving traditional artisan dairy products struggle to remain viable in competitive markets characterised by industrial production and weak market positioning. This study examines the viability of the SFSC for queso tenate, a traditional artisan cheese from central Mexico, through a cybernetic perspective using the Viable System Model (VSM) and the Viplan method. Accordingly, an integrative framework is proposed that combines cybernetic organisational design with a value chain perspective and a core infrastructure of food technology practices. The SFSC is analysed through the focal enterprise as the primary coordination and integration point for production, coordination, control, intelligence, and governance functions. The analysis incorporates technical and managerial activities, including food technology practices, production operations, and market-related processes. Using the Viplan method, the study represents systemic functions within the SFSC. The results identify structural weaknesses affecting viability, including fragmented coordination, limited technological validation, and insufficient market differentiation. The findings suggest that the configuration of systemic functions, as defined by the VSM, may be associated with organisational conditions shaping system functioning in traditional artisan food systems. The proposed framework provides a structured basis for diagnosing areas where viability may be strengthened. Limitations are acknowledged regarding the conceptual approach, the single-case study design, and the generalisability of the results. Future research may extend this work by examining diverse traditional cheese supply chains, exploring viability across multiple recursion levels, strengthening core infrastructure and market development activities, and incorporating stakeholders’ perspectives within SFSCs. Full article

Accurate state of charge (SOC) estimation is essential for reliable battery management systems operating over a wide temperature range. This study proposes a joint estimation framework that combines a temperature-matched dual open-circuit-voltage (OCV)-SOC model, online forgetting-factor recursive least squares (FFRLS), and an adaptive improved multi-innovation unscented Kalman filter (AIMIUKF). The dual OCV-SOC model separately calibrates charging and discharging branches at 0 °C, 25 °C, and 45 °C, reducing the voltage bias caused by thermal dependence and charge–discharge hysteresis. On this corrected voltage baseline, FFRLS identifies the time-varying parameters of the second-order RC equivalent circuit model. The updated parameters are then used by AIMIUKF, where a finite multi-innovation window improves convergence under initial SOC deviation, and covariance matching adjusts process and measurement noise online. Validation on the CALCE 18650 dataset under the Dynamic Stress Test (DST) profile shows sub-1% SOC errors at all tested temperatures. Full article

This paper introduces the Touchard process, a flexible two-parameter stochastic framework for modeling count data that depart from the classical Poisson assumptions. In contrast to standard Poisson processes, the proposed model allows for both nonstationary and dependent increments, enabling the representation of overdispersion, underdispersion, and temporal dependence within a unified structure. The main contribution lies in extending weighted Poisson models to a stochastic-process setting through recursively defined transition probabilities associated with Touchard marginal distributions. We derive key theoretical properties, including admissibility conditions and a recursive formulation for the transition probabilities, and propose an efficient simulation algorithm. Maximum likelihood estimation is developed for parameter inference, and a likelihood ratio framework is used for model comparison. An empirical application to motor vehicle crash data illustrates the ability of the model to capture dynamic patterns that are not adequately described by classical Poisson-based approaches. Full article

Accurate precipitation nowcasting is critical for many aspects of human life. A recurrent neural network (RNN) has demonstrated strong and relatively mature performance in machine learning approaches for precipitation nowcasting. However, their inherent recursive prediction structure leads to error accumulation, causing progressively blurred outputs and limiting practical applicability. To address this issue, we propose CSPP-RNN (Coupling-Similar-Precipitation-Processes RNN), a net that couples similar precipitation processes with a sequence-to-sequence RNN. For each prediction timestep, similar precipitation processes are retrieved, and their segments are then input into the encoder to obtain the corresponding hidden states. These hidden states replace the ones influenced by earlier predicted results in the recursive structure. Based on radar data from Beijing Daxing station, the comparison experiments of CSPP-RNN and ConvLSTM indicate that: (1) Over the 36–60 min lead time across the 0.1, 5.0, and 20.0 mm/h thresholds, the POD and CSI improved by 0.0334, 0.0170 on average, respectively, whereas the FAR degraded by 0.0586; (2) error accumulation was mitigated, retaining richer fine-scale structures in the predicted images; (3) the extra computational cost of coupling was controlled within an acceptable range. In conclusion, CSPP-RNN mitigates the error accumulation problem in RNN by coupling similar precipitation processes as part of the modification of the recursive prediction structure. This provides a potential new direction for optimizing the application of RNN in precipitation nowcasting. Full article

Existing anomalous traffic detection methods based on feature fusion in Software-Defined Networking (SDN) lack adaptability in weight allocation mechanisms. Consequently, their detection accuracy and model generalization capabilities fail to meet practical security requirements. To solve these limitations, this paper proposes a refined detection method based on hybrid feature selection and gated fusion. First, the framework employs XGBoost combined with the Recursive Feature Elimination (RFE) algorithm. This process identifies shallow statistical features with high discriminative power. Simultaneously, the method utilizes a 1D Convolutional Neural Network (1D-CNN) integrated with a Squeeze-and-Excitation (SE) block to extract deep temporal semantic features. Subsequently, a tailored gated fusion mechanism incorporating linear projection layers for feature alignment adaptively integrates these two categories of features. The fused features are then input into a Multilayer Perceptron (MLP) to execute anomalous traffic detection. Experimental results demonstrate that the proposed method achieves superior performance. Specifically, on the InSDN Dataset, the binary and multi-classification accuracy rates reach 99.91% and 99.88%. Similarly, the accuracy rates on the NSL-KDD dataset are 99.78% and 99.76%. Finally, we established a local simulation environment. Experimental results demonstrate that our method attains an average precision exceeding 93% for anomalous traffic detection in simulated real scenarios. Full article

To address the challenges of high computational complexity, inferior path performance, and the balance between path quality and efficiency in traditional 3D omnidirectional path planning algorithms for UAVs, this study proposes an innovative precision algorithm for solving 3D omnidirectional shortest paths. The algorithm innovatively introduces the concepts of circling path and overpass path, reducing three-dimensional omnidirectional path computation to two-dimensional processing. It designs a three-view obstacle detection algorithm to achieve efficient obstacle avoidance judgment, formulates separate path-solving strategies for discrete and continuous obstacles, respectively, and obtains optimal solutions through recursive adjustments and path optimization. Experimental results demonstrate that compared to A* and Theta* algorithms, our approach achieves shorter path lengths with superior stability; the proposed algorithm achieves a 21.86% reduction compared to RRT*, 10.48% compared to A*, and 0.89% compared to Lazy_Theta*. In addition, the proposed algorithm exhibits enhanced adaptability in high-obstacle environments (particularly irregular obstacles). These findings provide an effective solution for complex spatial path planning in UAV applications. Full article

Object detection in Unmanned Aerial Vehicle (UAV) imagery faces significant challenges due to the unique aerial perspective. A major bottleneck is the weak feature representation of small objects, which limits both detection accuracy and computational efficiency. To address this issue, we propose a Lightweight Detail-Sensitive Network (LDSNet). Specifically, LDSNet consists of three key components: (1) Lightweight Detail-Sensitive Downsampling (LDSDown), which combines anti-aliasing smoothing with dual-path feature extraction to preserve the spatial details of small objects during downsampling; (2) Shared Recursive Dilated Convolution (SRDC), which uses weight-shared multi-rate dilated convolutions to capture multi-scale context and enlarge the receptive field without introducing extra parameters; and (3) Deeply Decoupled Grouped Head (DGHead), which employs high-ratio grouped convolutions to significantly reduce the computational cost of processing high-resolution inputs. Extensive experiments on the VisDrone2019 and HIT-UAV datasets demonstrate that LDSNet achieves an excellent trade-off between accuracy and efficiency. Compared to the YOLOv11n baseline, LDSNet reduces parameters by 84.6% (from 2.6 M to 0.4 M) and FLOPs by 29.2% (from 6.5 G to 4.6 G), while improving mAP₅₀ by 2.2% on VisDrone2019 and achieving 94.5% on HIT-UAV. Full article

Recursive least-squares (RLS) adaptive algorithms are capable of outperforming least-mean-square (LMS) methods for the identification of long-length impulse responses due to their ability to mitigate the high correlation properties of input signals, such as speech. Despite the encouraging results obtained in terms of tracking speed and accuracy, with respect to LMS methods, most RLS algorithms manifest numerical stability issues. Moreover, when an unknown system changes, the identification process needs to adapt to the new impulse response as soon as possible. The algorithm can require a significant amount of time to generate new accurate results in acoustic echo cancellation (AEC) scenarios. Due to the slow propagation speed of sound, acoustic echo paths are usually modeled using thousands of numerical coefficients, and adaptation energy remains relatively limited. A compromise is usually made between tracking capabilities and steady-state accuracy when choosing the forgetting factor (the most important parameter of the RLS algorithm). This paper analyzes a variable forgetting factor (VFF) RLS type of adaptive filter combined with the conjugate gradient (CG) line search method, which is designed to avoid the classical matrix inversion approach. This VFF-RLS-CG adaptive method is not susceptible to numerical stability issues and is designed to adapt its statistical estimates by determining whether a tracking situation occurs or whether the unknown system is not significantly different. Correspondingly, when necessary, the forgetting factor is decreased for faster adaptation to changes in the working environment. When the filter is estimated to work at steady-state, the above-mentioned parameter’s value is increased in order to boost the accuracy of the adaptive filter. The theoretical model is validated using simulations in AEC scenarios with tracking occurrences and relevant steady-state intervals. Full article

Education for Sustainable Development (ESD) increasingly relies on e-learning environments to deliver scalable, flexible, and accessible educational experiences. While existing research has extensively examined the role of digital technologies in facilitating sustainability-oriented learning, significantly less attention has been given to how learners themselves adapt to the structures and dynamics of these environments. This study adopts a theory-building approach, grounded in integrative conceptual synthesis, to develop a coherent explanatory framework that combines insights from cognitive load theory, self-regulated learning, neurocognitive adaptation, and systems theory. The paper proposes the Sustainable Learning Loop, a recursive mechanism describing how repeated interaction with e-learning systems leads to pattern recognition, expectation stabilization, reduced cognitive effort, decreased active engagement, and the emergence of stable behavioral learning patterns. In addition, a novel typology of Sustainable Learning Habituation (SLH) is developed, distinguishing between cognitive, engagement, behavioral, and sustainability habituation. The framework highlights a dual dynamic in digital learning environments. On the one hand, habituation enhances efficiency, usability, and scalability, supporting the broader goals of sustainable education. On the other hand, it introduces critical risks, including superficial sustainability learning, reduced critical thinking, platform dependency, and the emergence of an illusion of learning, where perceived competence exceeds actual understanding. The study contributes to the literature by reframing e-learning as a co-adaptive system that actively shapes cognitive and behavioral processes, rather than merely delivering content. From a practical perspective, the findings underscore the need to design digital learning environments that balance efficiency with reflexivity, ensuring that ESD remains transformative rather than procedural. The framework also offers actionable design implications, suggesting that e-learning environments for ESD should incorporate reflective prompts, task variability, learning analytics indicators, and metacognitive feedback mechanisms to prevent excessive routinization and support deeper sustainability-oriented engagement. Future research should focus on the empirical operationalization of SLH and the development of adaptive systems that support critical engagement and sustainable learning outcomes. Full article