Search Results (689)

Accurate state of charge (SoC) estimation is of primary importance in terms of safe and efficient management of energy storage systems (ESSs). In this regard, data-driven frameworks offer the advantage of rapid execution during online operations. Nevertheless, their deployment on resource-constrained embedded systems is often hindered by the strict memory and processing limitations of low-cost hardware. This article proposes a three-stage pipelined SoC estimation framework incorporating a reinforcement learning (RL) primary stage, least squares boosting (LSB) secondary residual corrector, and ultimate linear three-point interpolation (3p-InT) stage. The RL phase utilizes a twin deep delayed deterministic policy gradient neural network (TD3NN)-based agent along with a customized reward function. The inference part of the framework is deployed on two separate embedded platforms, i.e., an STM32F411RE microcontroller (MCU) and Digilent Nexys A7-100T FPGA through automatic C code and hardware description language (HDL) code generation features in MATLAB/Simulink, respectively. The efficacy of the proposed framework is evaluated using a Panasonic 18650 lithium-ion battery (LiB) and a battery-powered drill load profile (BPD-LP). Across the four hardware scenarios, the accuracy of the proposed framework is preserved, with the maximum %RMSE deviation not exceeding 0.08 percentage points. The RMSE value remains within 1.80–1.82% for the LiB dataset and within 0.76–0.84% for the BPD-LP, irrespective of the platform or the arithmetic format. As for the resource footprint, the fixed-point implementation more than halves the FPGA logic with respect to the floating point (27.33% against 65.55% of the LUTs), at the cost of a comparatively higher DSP usage (15% against 7.08%). On the MCU, it trades additional flash memory (31.51% against 25.23%) for a 2.6-fold smaller RAM footprint. The framework’s reward function, hyperparameters, and architecture are kept unchanged across both datasets, indicating that the same configuration can be generalized across both profiles without requiring dataset-specific re-tuning. Moreover, the detailed hardware deployment findings provide a practical insight into the hardware and arithmetic format selection for an accurate embedded SoC estimation framework. Full article

To address time-varying communication topology, actuator response delay, and limited inter-UAV communication resources in the multi-UAV approach of a maneuvering target, this paper proposes a cooperative rendezvous/tracking control law combining a prescribed-time extended state observer (PTESO) with a dynamic event-triggered mechanism (DET). A three-state PTESO is designed whose observation error converges, within a prescribed time independent of the initial error, into a compact set related to the disturbance upper bound. Along the line-of-sight (LOS) direction, the remaining flight times of the UAVs are driven to consensus within a prescribed time toward a specified arrival instant via a threshold-adaptive DET; along the LOS normal direction, prescribed-time convergence of the elevation and azimuth angle errors is achieved through a time-varying-gain sliding surface. The guidance gains are designed from the worst-case algebraic connectivity of the candidate topology set, ensuring uniform validity under arbitrary switching. After accounting for first-order autopilot inertial dynamics, the command tracking error is proven uniformly ultimately bounded. Simulations of four UAVs cooperatively approaching a maneuvering non-cooperative object under periodic topology switching and actuator delay show an arrival-time deviation below 0.01 s, a terminal position error under 0.08 m, and 75–91 average inter-UAV triggers, outperforming existing prescribed-time/fixed-time cooperative guidance methods. Full article

Formations of fixed-wing unmanned aerial vehicles (UAVs) must maintain mission-level coordination despite actuator degradation, bias faults, model uncertainty, wind disturbances, and command limits. This study presents a mission-oriented formation-recovery framework. Acceptable relative position and velocity performance is represented by an ellipsoidal tolerance set, and resilient formation recovery time (RFRT) is the elapsed time from fault onset to the first permanent re-entry into that set. The controller generates relative demand commands that are reconstructed from the parent limited command and then clipped componentwise. A second-order edge-based extended state observer, driven by the relative limited command, estimates the physical edge fault–disturbance mismatch, while the command-saturation residual is retained explicitly in the closed-loop analysis. A distributed fault-tolerant controller combines linear recovery feedback, observer-based compensation, and a continuous robust term. The analysis uses a finite Lyapunov energy-jump inequality at step-fault instants and an upper-right Dini derivative at saturation breakpoints to establish uniform ultimate boundedness, a sufficient recovery-set inclusion condition, and an RFRT upper bound. In the one-leader–four-follower baseline simulation, permanent re-entry occurs at 22.3313 s, corresponding to an RFRT of 2.3313 s. Relative to Edge-ESO-FTC, the proposed method reduces the post-fault peak metric by 61.2%, RFRT by 25.2%, cumulative position deviation by 54.6%, and out-of-bounds duration by 67.7%. Full article

Prophages are widespread in bacterial genomes and can influence pathogen evolution by modulating virulence, antimicrobial resistance, environmental adaptation, and competitive fitness. However, prophage research in aquaculture pathogens remains limited, with abundant genomic predictions but scarce functional validation, numerous cross-sectional surveys but little longitudinal tracking, and extensive phenomenological observations but insufficient mechanistic investigation. This review synthesizes current knowledge of prophages in aquatic bacterial pathogens, focusing on their roles in virulence, antimicrobial resistance, and competitive advantage, while drawing on mechanistic paradigms established in human and animal pathogens. We further propose a pathogen-centered “molecular bridge” framework in which environmental perturbations alter bacterial physiological states, are interpreted through prophage regulatory mechanisms, and ultimately translate into changes in pathogen phenotypes, fitness, and pathogenic potential. Particular attention is given to multi-signal interactions, induction thresholds, and the context-dependent fitness consequences of prophage induction in dynamic aquaculture environments. Finally, we propose future priorities encompassing systematic prophage resource construction, functional validation, multi-stressor experiments, and longitudinal monitoring. By adopting a paradigm-borrowing review approach, this review aims to provide a transferable roadmap for prophage research in aquaculture, to facilitate translational applications, and to offer theoretical support for precision disease control in aquaculture. Full article

Advanced analysis has been shown to improve material efficiency in statically indeterminate steel-framed structures compared with member-based linear elastic design methods. However, limited research has investigated its applicability to geometrically nonlinear steel structures where residual stresses are induced by the bending process. In this study, the material optimization potential of advanced analysis has been quantified for two arch-based structures by comparing the volume of steel required to satisfy the criteria of both the system and member-based analysis methods in accordance with AS 4100:2020. The two structures were analyzed using the finite element analysis software Strand7 (R3.1.6) and subjected to combined gravity and wind loading in alignment with the serviceability and ultimate limit states specified in AS 1170.0:2002. System behavior was analyzed through the Arc-length plastic zone method. The results indicate that in one of the arch-based structures, advanced analysis can improve material utilization by 8.1%. Provided that future research both validates the use of the reduced stiffness method for treatment of initial geometric imperfections and verifies system reliability factors for structures with curved geometries, advanced analysis presents a practical design method for this structure. Comparison of the two case studies found that advanced analysis has the potential to improve material efficiency only when linear elastic failure is governed by ultimate limit state criteria. It is therefore evident that the material optimization findings of this research cannot be generalized to all arch-based structures, as they are contingent upon the geometry of the model analyzed, the loading scenarios considered, and the deflection limits adopted. Full article

Despite the growing use of FLNG in offshore gas development, LNG hose safety during tandem offloading remains a critical challenge. Existing studies often analyze mooring dynamics and hose mechanics separately, lacking a unified framework that integrates multiple failure modes. This fragmented approach leads to unclear safety boundaries and inadequate risk control. Therefore, this study proposes a multi-parameter safe operating zone threshold method based on coupled dynamic analysis. First, a three-dimensional time-domain dynamic analysis model is developed using OrcaFlex, which integrates the floating bodies, hoses, and mooring system into a unified coupling framework based on hydrodynamic theory, simulating the dynamic response of the offloading system under combined wind, wave, and current actions. Second, tension, bending moment, and curvature are selected as safety evaluation parameters. These three parameters correspond to the core criteria of typical failure modes, namely axial overload failure, ultimate bending failure, and local joint failure, respectively. By comparing them with their allowable values, the safety status of the hose under various operating conditions is determined. Finally, a coupled safety threshold analysis method incorporating both “sea state return period” and “operational vessel distance” is proposed. The results indicate that, at a fixed vessel distance, the dynamic response of the hose increases significantly with worsening sea states. Tension satisfies the safety factor requirements under most sea conditions. However, the bending moment first exceeds the limit starting from the 5-year return period, making it the primary failure control indicator. Curvature exceeds the limit notably under the 50-year return period and beyond, becoming the main risk source under extreme sea states. The safe operational vessel distances under different sea states are also calculated, systematically revealing the response patterns and failure sequences of tension, curvature, and bending moment of the LNG hose under combined wind, wave, and current actions. Furthermore, by integrating safety margin calculations, an operational classification standard comprising a safe zone, a warning zone, and a danger zone is proposed, along with the upper limits of safe vessel distance and operational windows for each sea state. The threshold determination method established in this paper can provide effective engineering support for FLNG offloading operation planning, hose selection, and operational risk management. Full article

Holes in the flanges of steel beams and girders are commonly required for bolted connections and can influence both flexural strength and rotational ductility. Although flange-hole effects have been investigated previously, the experimental basis of several traditional provisions was developed using older steels with relatively low yield-to-ultimate strength ratios. Current design specifications differ considerably in their treatment of open and fastener-filled flange holes. In particular, the relationship between flange-area reduction, flexural resistance, rotational ductility, and net-section fracture in modern structural steels remains insufficiently quantified. This study addresses these issues through tests of twenty-five full-scale W200 × 42 beams fabricated from ASTM A992 steel with a measured (F_y/F_u) ratio of approximately 0.77. The test program included solid beams, beams with open holes in the tension flange, beams with open holes in both flanges, and beams with fastener-filled holes in both flanges. The results demonstrate that flange-area reduction affected rotational ductility more severely than flexural strength. For the most severe tension-flange reduction investigated (A_fn/A_fg = 0.52), the maximum moment decreased by 16.8% relative to the solid beams’average maximum moment, whereas the total rotation capacity (R_y) decreased by 77.2%. The experimental capacities were also compared with predictions from AISC-LRFD, AASHTO-LRFD, CAN/CSA-S16, and AS 4100. Based on the observed strength and failure behavior, a design framework considering gross-section flexural resistance and modified net-section fracture resistance as competing limit states is proposed. Across the investigated specimens, the ratio of measured maximum moment to the proposed design moment (M_m/M_dp) ranged from 1.21 to 1.47, indicating conservative predictions within the experimental range considered. Full article

Maglev trains are susceptible to various types of operational challenges, including track irregularities, load variations, and actuator faults. It is evident that these factors can compromise suspension performance and even pose a serious risk to operational safety. This paper proposes a prescribed performance event-triggered fault-tolerant control method for the electromagnetic suspension system of a maglev train subject to multiple constraints. A projection-based adaptive extended state observer is designed to estimate the unknown gain caused by actuator faults and load variations, as well as the external disturbance. In light of the disparity in upper and lower safety margins inherent to the suspension gap error, arising from track irregularities, an asymmetric prescribed performance function and an error transformation are devised to ensure that the gap tracking error perpetually complies with the asymmetric prescribed performance constraint. In addressing the issue of rapid variations in the suspension gap, the vertical velocity is also constrained through the implementation of prescribed performance, resulting in a joint constraint framework that encompasses both the gap tracking error and the vertical motion. A dynamic event-triggered mechanism has been incorporated into the backstepping design with a view to reducing unnecessary control updates under limited communication resources, while Zeno behavior has been excluded from the closed-loop system. Within this framework, a dynamic gain adjustment mechanism with an explicitly bounded rate of variation is further developed to achieve smoother gain adaptation. The uniform ultimate boundedness of all closed-loop signals is demonstrated through Lyapunov stability analysis under the prescribed multiple constraints. The efficacy of the proposed method is demonstrated through comparative simulation results. Full article

The exponential growth of the Internet of Things (IoT) has generated massive data streams traditionally processed by centralized cloud architectures, which increasingly face latency, bandwidth, and privacy limitations. Shifting artificial intelligence to resource-constrained edge nodes, known as TinyML, offers a robust decentralized alternative, though it introduces severe memory, compute, and energy bottlenecks. To map this transition, a systematic literature review was conducted following PRISMA guidelines, analyzing peer-reviewed studies published between 2021 and 2026 across major databases. The analysis identifies primary architectural paradigms and evaluates the efficacy of state-of-the-art model compression techniques, such as quantization, pruning, and knowledge distillation. Furthermore, the findings reveal that hardware–software co-design and custom neural accelerators are crucial for overcoming operational bottlenecks, while also highlighting persistent security and privacy challenges in on-device learning. Ultimately, while deploying complex models on microcontrollers is increasingly viable, achieving optimal performance demands holistic optimization strategies. This review synthesizes current research gaps and provides a strategic roadmap to guide future interdisciplinary efforts toward resilient, energy-efficient, and secure next-generation intelligent edge systems. Full article

Upper-limb motor dysfunction resulting from neurological disorders severely limits patients’ activities of daily living and social participation. Pneumatic upper-limb rehabilitation robots have emerged as a promising intervention owing to their inherent compliance, lightweight design, and high power-to-weight ratio, which facilitate safe, repetitive, and home-based training. Despite these advantages, extensive clinical translation remains hindered by challenges including actuator hysteresis, nonlinear dynamics, limited accuracy in intention recognition, and inconsistent clinical evaluation metrics. This review systematically examines recent advancements in pneumatic upper-limb rehabilitation robots across four critical dimensions: structural design, human–robot interaction, control strategies, and clinical translation. We comparatively analyze rigid exoskeletons, soft wearable devices, and rigid–soft hybrid configurations based on output capability, motion accuracy, comfort, and clinical applicability. The findings suggest that while rigid systems offer high precision and soft systems maximize safety, rigid–soft hybrid architectures represent a critical developmental trend for balancing motion accuracy with interaction compliance. Furthermore, the review evaluates multimodal sensing techniques (e.g., EMG, EEG, and IMUs) for motion intention decoding and training state monitoring, alongside conventional, adaptive, and artificial intelligence-driven control methods aimed at compensating for pneumatic nonlinearity and improving real-time response. Current clinical evidence indicates that these systems effectively enhance upper-limb function and muscle strength, particularly in post-stroke rehabilitation; however, existing trials are frequently constrained by small sample sizes, short interventions, and heterogeneous protocols. Future research must prioritize rigid–soft hybrid architectures, robust multimodal sensor fusion, digital twin-assisted assessment, adaptive intelligent control, and standardized home-based rehabilitation platforms. Ultimately, this comprehensive review provides a concise reference for the design optimization and clinical deployment of next-generation pneumatic rehabilitation systems. Full article

The evaluation of the ultimate bearing capacity (UBC) of foundation systems remains a fundamental challenge in geotechnical engineering because of the complex interactions among soil properties, foundation geometry, loading conditions, embedment depth, and soil–foundation behavior. In recent years, artificial intelligence (AI) and deep learning (DL) techniques have emerged as powerful data-driven tools for modeling nonlinear geotechnical systems and improving bearing-capacity prediction. This study presents a comprehensive state-of-the-art review of AI- and DL-based approaches for foundation systems, including shallow foundations, deep foundations, pile foundations, and other geotechnical applications. Major models, including Artificial Neural Networks (ANNs), Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks, Transformer models, Graph Neural Networks (GNNs), hybrid AI frameworks, and physics-informed deep learning approaches, are critically reviewed and compared. Particular attention is given to the integration of AI models with numerical methods, including the finite element method (FEM) and finite element limit analysis (FELA). The reviewed studies frequently report lower prediction errors than conventional empirical, numerical, and machine-learning approaches within the evaluated datasets. However, many of the highest reported accuracies are based on laboratory-scale experiments, simulation-generated data, or random train–test partitions of a single database. Consequently, these results may demonstrate effective interpolation within controlled data distributions rather than reliable performance under independent field conditions. Model performance is strongly influenced by dataset origin and diversity, feature selection, validation strategy, overfitting control, and generalization capability. Hybrid datasets combining field, laboratory, and numerical data offer a promising route toward more reliable prediction, but genuine external validation using independent sites, projects, or institutions remains uncommon. Moreover, architectural suitability should reflect the physical structure of the problem: CNNs are appropriate for spatial heterogeneity, LSTMs for time-dependent behavior, Transformers for long-range interactions, and GNNs for mechanically connected systems. Limited field-scale datasets, weak external validation, limited model interpretability, inadequate uncertainty quantification, and persistent data scarcity continue to restrict widespread engineering implementation. Future research should prioritize explainable AI, physics-informed learning, transfer learning, hybrid data frameworks, open benchmark datasets, and multi-site field validation to improve the robustness, transparency, and practical applicability of intelligent bearing-capacity prediction for diverse foundation systems. Full article

To addressthe challenge of attitude control in quad-ducted fan unmanned aerial vehicles (UAVs) under coupled disturbances comprising thrust lag, ground effect and a composite wind field during near-ground flight and to mitigate the inherent trade-off between disturbance rejection and noise suppression in fixed-bandwidth extended state observers (ESOs), this paper proposes a robust attitude control method based on a Sigmoid law adaptive-bandwidth extended state observer (AB-ESO). An attitude dynamic model covering the above multi-source disturbances is established, with all uncertainties uniformly treated as lumped disturbances. An adaptive-bandwidth mechanism with filtering and rate-limiting modules is designed for smooth continuous bandwidth tuning. A composite control framework integrating disturbance feedforward, lag compensation and attitude feedback is constructed, and the uniform ultimate boundedness of the closed-loop system is proved. Comparative simulations are conducted against six baseline controllers, including a cascade proportional–integral–derivative (PID) controller, fixed-bandwidth ESOs, incremental nonlinear dynamic inversion (INDI), fast terminal sliding mode control (FTSMC) and a time-varying bandwidth ESO, in a near-ground composite wind scenario. Results show that the proposed method achieves improved comprehensive performance: the three-axis average tracking root mean square error (RMSE) is approximately 72% lower than of the PID controller and 15.8% lower than that of the high-bandwidth ESO, and the control output total variation is reduced by about 27.8%. Monte Carlo verification with 100 random turbulence groups further validates the strong statistical robustness of the proposed method. All validations in this work are based on numerical simulations. This study provides a technical reference for high-precision control of ducted fan UAVs in near-ground environments. Full article

Electroencephalography (EEG) measured inside or around ears, called ear-EEG, provides a practical measurement modality for daily brain–computer interface (BCI) applications. However, reliable decoding of mental imagery remains challenging due to the limited number of channels, low signal-to-noise ratio (SNR), and substantial inter- and intra-subject variability inherent to ear-EEG. Addressing these constraints requires advanced decoding strategies specifically optimized for this signal domain. In this study, we retrospectively analyze the ear-EEG dataset of a previous study in which a real-time endogenous BCI was evaluated using conventional machine learning. Specifically, we present an offline benchmark of 23 deep neural network architectures originally developed for scalp-EEG, which were adapted to ear-EEG and evaluated under an identical validation framework. To the best of our knowledge, this is the first systematic comparison of this breadth for ear-EEG-based mental-task classification. Beyond conventional performance comparison, we identify the optimal architecture by jointly considering statistical significance and a performance–cost trade-off, incorporating classification accuracy, parameter count, and measured computational cost. Our results demonstrate that FBLightConvNet achieves the highest classification accuracy among all evaluated models and outperforms common spatial pattern-linear discriminant analysis (CSP-LDA), a widely adopted and robust conventional baseline, on all three recording days, with the difference reaching statistical significance on Days 2 and 3. Notably, many state-of-the-art scalp-EEG models fail to generalize effectively to ear-EEG, highlighting the importance of architecture selection in this domain. These findings identify the best-performing architecture in this setting and indicate which architectural characteristics support effective ear-EEG decoding. Ultimately, this study offers practical design insights and a reproducible benchmarking framework for developing lightweight and high-performance deep learning models, which we hope will support future efforts toward real-world ear-EEG-based BCI systems. Full article

The replacement of synthetic dyes with natural food colorants has become a priority for the food industry, as emerging evidence from in vitro and animal studies on the potential neurotoxic and pro-inflammatory effects of certified dyes converges with consumer pressure for clean-label formulations. Yet despite this, the industrial uptake of natural pigments remains uneven, held back by stability limitations that differ considerably from one pigment class to the next and from one food matrix to another. This review covers the chemistry, industrial applications, and stabilization approaches of the main natural colorant groups: carotenoids, anthocyanins, betalains, chlorophylls, curcuminoids, phycocyanin, and genipin-derived pigments, with particular attention to the physicochemical reasons behind their instability and the practical tools available to address it. Among stabilization strategies, spray-drying microencapsulation with composite protein–polysaccharide wall materials is often the most scalable and cost-effective option, whereas freeze drying may be preferable for high-value pigments or applications in which maximum pigment retention is the priority. Whether the encapsulating matrix remains in a glassy or rubbery state stands out as a key factor governing oxidative degradation across all pigment categories, which makes water activity management a non-negotiable element of any serious formulation effort. Anthocyanins require more than physical encapsulation alone: copigmentation and structural approaches such as acylation and pyranoanthocyanin formation hold degradation routes that no shell material can prevent on its own. For hydrophobic pigments like carotenoids and curcuminoids, lipid-based delivery systems consistently deliver higher bioaccessibility than aqueous or dried formats. pH control, antioxidant incorporation, and modified atmosphere packaging add a useful but ultimately incomplete third line of defense. One development worth attention is the use of pH-responsive pigments in biopolymer packaging films, where color instability, long treated as a drawback, becomes a real-time indicator of food freshness. Bridging the remaining performance gap with synthetic dyes will call for stabilization platforms that tackle the molecular, physical, and environmental dimensions of degradation together, built around the particular chemistry of each pigment and the demands of each application. Full article

ViTs have achieved excellent performance in image recognition tasks, but their large parameter counts and high computational complexity limit their deployment on resource-constrained devices. Most existing ViT pruning methods adopt class-agnostic pruning strategies, which fail to distinguish the diverse structural requirements of different target classes. As a result, they are prone to removing critical features, leading to class-wise accuracy imbalance in practical deployment. To address this issue, this paper proposes a class-aware joint pruning framework for ViTs, which collaboratively compresses the model from two orthogonal dimensions: MLP neurons and visual tokens. Specifically, (1) based on first-order Taylor expansion, we quantify the contribution of each MLP neuron to the target classes and adaptively prune redundant neurons to achieve structured compression, followed by lightweight fine-tuning on the target class subset; (2) we propose a Class-Guided Token Selection (CGTS) method, which constructs class prototype vectors using a few support samples of the target classes and then dynamically selects patch tokens that are semantically highly relevant to the target classes during inference in a zero-shot manner, requiring no additional training or fine-tuning. The two modules complement each other, achieving dual compression from the parameter dimension and the inference data dimension. Experiments on CIFAR-100 and TinyImageNet datasets using DeiT-Tiny/Small models demonstrate that, compared with state-of-the-art pruning methods, our method reduces GMACs on target class subsets by up to 48%, improves inference speed by nearly 50%, and requires only 0.8 KB of additional storage overhead per subset, ultimately achieving a superior trade-off among accuracy, computational efficiency, and storage overhead. Full article