Search Results (2,038)

Accurate state-of-charge (SOC) estimation is essential for battery management systems (BMSs) under dynamic operating conditions. This study proposes a lightweight data-driven SOC estimation framework based on a tensor-product bivariate polynomial surface. During offline model identification, the reference SOC and measured current are used to establish an explicit terminal-voltage surface (V = f(I, SOC)). During SOC estimation, however, the reference SOC is no longer used as an input; instead, SOC is directly recovered from the measured terminal voltage and current by inversion of the identified polynomial surface. The method used in this study is developed and evaluated using three public dynamic driving-cycle datasets, including DST, FUDS, and UDDS. To reduce temporal information leakage, the datasets are partitioned chronologically, and the polynomial orders are selected using five-fold forward-chaining cross-validation. The selected third-order current and fifth-order SOC polynomials provide a compact 24-coefficient representation. Direct comparison between the estimated and reference SOC yields an RMSE of 2.91% on the DST test set, while the corresponding RMSE values on FUDS and UDDS are 3.18% and 3.54%, respectively. The results demonstrate that the proposed approach provides a computationally lightweight and interpretable SOC estimation framework without requiring recursive filtering or online equivalent-circuit parameter identification. Full article

The Internet of Vehicles (IoV) relies on frequent vehicle-to-roadside-unit (RSU) access over open wireless channels, making efficient authentication and key establishment essential. In many existing authentication and key agreement (AKA) schemes, a target RSU receives and processes a request before an invalid sender is rejected, which can waste roadside computation under dense invalid-request traffic. This paper presents EDAKA-IoV, an elliptic-curve-cryptography-based AKA scheme that separates admission filtering from end-to-end session-key establishment. A trusted authority (TA) performs Lightweight Polynomial-based Pre-Verification (LPPV) to discard invalid authentication requests before they reach the target RSU, while the vehicle and RSU establish the final session key. A current–pending dual-state mechanism prevents permanent de-synchronization during dynamic pseudo-identity renewal without adding another communication round. Formal analysis under an eCK-style model, ProVerif verification, and heuristic analysis evaluate session-key secrecy, injective mutual authentication, privacy, and resistance to the considered attacks. Optional offline precomputation moves two fixed-base scalar multiplications outside the online phase and reduces the non-polynomial online computation component by approximately 48.8%. With fixed-length compressed point encoding, the authentication exchange requires 2336 bits. The workload analysis shows that the RSU-side processing reduction is proportional to the invalid-request ratio, while the TA still performs record lookup, hashing, and degree-dependent polynomial evaluation for every received request. EDAKA-IoV therefore provides a balanced authentication solution for resource-sensitive IoV deployments. Full article

Off-grid photovoltaic (PV) hydrogen production systems must coordinate rapidly varying PV power, battery energy, and the operating states of multiple alkaline water electrolyzers. Inappropriate coordination may lead to PV curtailment, frequent unit switching, and persistent workload concentration on a small number of electrolyzers. This paper develops an efficiency- and load-balanced operation (ELBO) scheme as an improved rule-based supervisory strategy rather than an online optimization method. ELBO adopts a two-level decision structure. A planned number of online electrolyzers is first determined from the moving-average PV power and the reference power associated with high single-unit efficiency. This planned count is then corrected using real-time PV power, battery state of charge, and the previous electrolyzer states. The controller adjusts the powers of the online units before changing their number, uses the battery to bridge temporary power deficits, and distributes the remaining adjustable power under the operating and ramp-rate constraints. Five representative PV profiles selected from one year of measured data were used to compare ELBO with PV-following operation (PFO), multi-electrolyzer coordinated operation (MECO), and an offline mixed-integer linear programming (MILP) benchmark. ELBO produced 1328 kg of hydrogen, which was 8.85% and 6.07% higher than PFO and MECO, respectively. Its overall PV-to-hydrogen efficiency and PV utilization reached 65.2% and 94.9%, respectively, with 36 start–stop events. MILP produced 1345 kg of hydrogen, only 1.28% more than ELBO, but required the complete future PV sequence. Ablation analysis further shows that the planned-count layer, moving-average filtering, battery-supported retention, and load-balancing allocation contribute to different and complementary aspects of capacity matching, operating continuity, and workload distribution. The results indicate that the benefit of ELBO arises from the ordered coordination of these supervisory functions and that it provides a practical compromise between operating performance, workload distribution, information requirements, and computational complexity under the representative conditions considered. Full article

The rapid advancement of software-defined networking (SDN) has enhanced network programmability, centralized control, and traffic management flexibility, while also increasing exposure to sophisticated attacks targeting the control plane. Although federated learning (FL) enables collaborative intrusion detection without centralized raw data sharing, existing FL-based intrusion detection systems remain vulnerable to plaintext model update leakage, centralized cryptographic trust, limited interpretability, and insufficient validation in operational SDN environments. To address these limitations, this paper presents an explainable federated intrusion detection framework that integrates distributed key generation (DKG), CKKS-based threshold homomorphic encryption, collaborative decryption, and SHapley Additive exPlanations (SHAP). Unlike conventional HE-enabled FL systems that rely on a trusted authority or a globally shared secret key, the proposed framework removes the trusted key-generation dealer, avoids centralized custody of the complete secret key, and prevents any single client or aggregation server from independently decrypting ciphertexts using locally held key material. A gated recurrent unit (GRU)-based model is used for privacy-preserving intrusion detection, and SHAP provides global and local explanations of model decisions. The framework is further deployed in a real-time SDN testbed to evaluate the online inference pipeline following threshold-secured federated training. Computationally intensive cryptographic operations, including DKG, encrypted aggregation, and threshold decryption, are performed during offline training, while the converged global model enables low-latency inference at runtime. Experiments on the InSDN, CICDDoS2017, and CICDDoS2019 datasets with 4, 8, and 12 client federated configurations achieved detection accuracies above 99% across all datasets. The evaluation also examines encryption latency, collaborative decryption overhead, secure aggregation cost, communication complexity, and scalability. The results demonstrate that the proposed framework provides a practical balance among decentralized key management, privacy-preserving aggregation, explainability, detection performance, and real-time SDN deployment feasibility. Full article

Global Positioning System (GPS) positioning accuracy is strongly affected by inaccurate noise modeling and non-Gaussian pseudorange measurement errors, including heavy-tailed disturbances and abnormal outliers caused by multipath propagation and signal degradation. Conventional extended Kalman filters (EKFs) generally assume Gaussian measurement noise with fixed covariance matrices, which limits their robustness under degraded measurement conditions. This study proposes a Student-t robust quantum-behaved particle swarm optimization-based extended Kalman filter (ST-QPSO-EKF) for adaptive GPS state estimation. The proposed framework combines quantum-behaved particle swarm optimization (QPSO) with a Student-t-based robust measurement update, where the process-noise scaling factor, measurement-noise scaling factor, and Student-t degrees-of-freedom parameter are jointly optimized. The optimized parameters are obtained through an offline calibration stage and subsequently applied in the recursive GPS filtering process. A nonlinear GPS navigation simulation was conducted using Gaussian, Student-t heavy-tailed, and outlier-contaminated pseudorange measurement scenarios. The proposed method was compared with conventional EKF, QPSO-EKF, and Student-t EKF using 20 independent Monte Carlo realizations. The results demonstrate that QPSO-EKF provides improved accuracy under nominal Gaussian conditions, whereas ST-QPSO-EKF achieves superior performance under non-Gaussian measurement environments. Under Student-t heavy-tailed noise, ST-QPSO-EKF reduced the position RMSE to 3.814 m, while under outlier-contaminated noise it achieved a position RMSE of 3.952 m, outperforming the other compared methods. In addition, the proposed method maintained comparable online computational cost because the QPSO optimization was performed offline. The results indicate that jointly optimizing covariance parameters and Student-t robustness provides an effective strategy for improving GPS positioning reliability under complex pseudorange measurement conditions. Full article

The rapid penetration of converter-interfaced renewable generation has reduced effective inertia in islanded microgrids, making frequency regulation increasingly sensitive to renewable intermittency, generation outages, and storage stress. This paper proposes CMSA-OVSG–EMCS, a coordinated dual-layer control architecture for islanded PV–diesel microgrids with battery–supercapacitor hybrid energy storage. Its main novelty lies in coupling an adaptive grid-forming CMSA-OVSG layer, which updates virtual inertia and damping online according to disturbance severity, with a supervisory EMCS that coordinates multi-time-scale HESS power sharing through the common DC link. In contrast to OVSG approaches that rely on offline tuning of fixed controller parameters, the proposed framework uses physics-constrained multi-scenario optimization to jointly account for frequency response, DC-link regulation, converter operating limits, and battery stress. Nonlinear simulations under load variations, renewable intermittency, PV disconnection, and diesel-generator outage show that the proposed method consistently delivers the strongest transient performance among the tested controllers. In the worst-case diesel-generator outage scenario, it reduces the maximum ROCOF by 51.5% and the battery-stress index by 27.5% relative to the strongest benchmark controller. Full article

Accurate lower-limb joint angle prediction can support motion-state perception and rehabilitation-oriented analysis. This offline feasibility study evaluated an FA-TA-BiLSTM model using surface electromyography (sEMG) features, together with historical hip and knee joint angles and angular velocities. Data were collected from five healthy adult male participants during level walking and sit-to-stand transitions, and the prediction horizon was 100 ms. Support vector regression (SVR), BiLSTM, and FA-TA-BiLSTM were compared under the same combined-input condition and evaluation protocol. The FA-TA-BiLSTM model achieved RMSE, MAE, and $R^2$ values of 2.0684°, 1.5920°, and 0.9726 for hip prediction and 2.9604°, 2.5142°, and 0.9660 for knee prediction, respectively. These values were obtained using a mixed-participant chronological split and should be interpreted as preliminary within-cohort estimates rather than evidence of participant-independent generalization. Under this restricted protocol, FA-TA-BiLSTM produced lower errors than SVR and standard BiLSTM. The participant-level model ranking was consistent across the five participants, but exact pairwise comparisons did not reach significance after Holm adjustment. The current comparison does not isolate the incremental contribution of sEMG or individual attention modules; larger and more diverse cohorts, participant-independent validation, modality and module ablation, and causal online evaluation remain necessary. Full article

EEG-based motor imagery (MI) decoding is commonly evaluated in open loop, but strong offline performance does not necessarily translate into stable real-time control once decoder outputs drive a physical device. This issue is especially relevant for lower-limb exoskeletons, where initiating movement, sustaining movement, stopping movement, and maintaining rest can impose different controller-level demands. This study presents a proof-of-concept controller-level validation of an EEG-driven ankle exoskeleton framework, using a staged design that links open-loop decoder development to real-time closed-loop controller testing. Open-loop EEG data were collected from nine able-bodied participants during static and dynamic ankle MI using an eight-channel g.tec Unicorn Hybrid Black system, with matched PA-SEMI outputs available for eight participants in the primary open-loop comparison. We used a hybrid feature representation combining spectral, spatial covariance, and temporal complexity descriptors to compare a supervised Passive–Aggressive (PA) classifier with a Semi-supervised latent learning network (SEMI). Performance was assessed using epoch-level and persistence-based event metrics intended to reflect controller triggering. Under the evaluated model-specific protocols, SEMI produced higher open-loop accuracy and lower false-trigger rates than PA. The reported SEMI analysis was transductive: feature windows from the target-participant evaluation runs were available without labels during consistency training, and their labels were withheld until final evaluation. However, we selected PA for the primary matched closed-loop validation because it could be retrained, checked, and deployed within the same-day workflow. Since SEMI was not evaluated in a balanced matched closed-loop comparison, this study does not determine whether PA or SEMI provides superior real-time controller performance. Deployment-oriented PA updates were audited using limited same-day calibration data and evaluated in matched PA-based closed-loop trials with three participants, based on online controller logs. The closed-loop experiments were conducted on RehAnkle, a pre-commercial robotic ankle rehabilitation device operated here as a single-active-DoF ankle platform for dorsiflexion-oriented EEG control. In closed-loop trials, start and stop commands were generally reliable, whereas sustained movement and sustained rest were less stable. These proof-of-concept results indicate that command generation and state maintenance should be evaluated as separate controller-level problems, and that open-loop accuracy alone is insufficient to characterize real-time exoskeleton control. Full article

Lightweight metro carbodies may exhibit elastic modes within ride-comfort-relevant frequency bands, limiting semi-active suspension controllers tuned offline for nominal conditions. This study proposes a skyhook-based parameter self-tuning fuzzy control (PSTFC) strategy for lateral secondary suspension. Its rule base and membership functions remain fixed, whereas two input quantization factors and one output scaling factor are updated online from the carbody lateral velocity and carbody–bogie relative lateral velocity, enabling state-dependent adaptation without increasing fuzzy-inference complexity. A rigid–flexible coupled multibody model is developed using Craig–Bampton component-mode synthesis and validated against field vibration measurements from a Type-A metro lead car. The controller is evaluated using ADAMS/Rail–MATLAB co-simulation, with robustness examined through repeated stochastic simulations and variations in vehicle speed, passenger load, track-irregularity intensity, and suspension parameters. The flexible model reproduces the measured location-dependent spectral characteristics more accurately than the rigid-carbody model. Under nominal conditions, PSTFC reduces the rear-carbody lateral-acceleration RMS from 0.1341 to 0.1015 m/s² and the Sperling ride comfort index from 1.5485 to 1.2864, corresponding to improvements of 24.3% and 16.9% over passive suspension. Relative to fixed-parameter fuzzy skyhook control, the two indicators are further reduced by 5.8% and 4.7%, respectively. The improvement persists across the investigated off-nominal conditions without controller retuning. These results demonstrate that state-dependent parameter scaling improves the adaptability of fuzzy skyhook control while retaining a compact inference structure, providing a computationally tractable approach to flexible-carbody vibration suppression. Full article

Low-emissivity (low-E) coatings are among the most effective thin-film technologies for reducing radiative heat losses and controlling solar heat gain in buildings, which account for approximately 30–40% of global primary energy consumption. This expanded review follows the technological evolution of low-E glazing from early transparent-conductor “heat mirrors” to modern multi-silver dielectric/metal/dielectric (D/M/D) architectures and emerging functional coatings. Four complementary perspectives are addressed: (i) the materials employed, from silver-based multilayers and transparent conducting oxides (ITO, FTO, AZO, GZO) to seed, blocker, and protective dielectric layers; (ii) the deposition systems, contrasting on-line pyrolytic/CVD “hard” coatings with off-line magnetron-sputtered “soft” coatings, together with ALD, sol–gel, and evaporation routes; (iii) the methods of analysis used to correlate microstructure, composition. and interfaces with optical, electrical, and thermal behaviour (XRD, XRR, SEM/TEM, AFM, XPS, SIMS, spectrophotometry, ellipsometry, emissivity, and U-value metrology according to EN 410/EN 673 and ISO 9050); and (iv) the functional performance of low-E stacks in insulating glass units, vacuum glazing, retrofit films, and smart-window systems across climate zones. Persistent research gaps are identified in long-term durability and ageing, indium-free scalable materials, standardized accelerated testing, and multi-objective design of thinner, more selective, and more robust stacks. Full article

The increasing penetration of distributed energy resources (DERs) introduces significant operational uncertainties, challenging the reliability and robustness of conventional protection coordination in distribution networks. Traditional optimization methods can obtain high-quality relay settings, but their iterative computational burden limits their direct use for real-time adaptation. To address this issue, this paper proposes a deep learning-based preventive relay setting calibration framework for enhancing protection coordination robustness in DER-rich distribution networks. The proposed method adopts an offline–online architecture. In the offline stage, a refined heuristic algorithm is integrated with ETAP-based fault analysis to generate a comprehensive dataset of high-quality optimized time-multiplier settings (TMSs) and pickup current settings (PCSs) under a wide range of DER-generation and load-demand scenarios. Subsequently, a convolutional neural network (CNN) is trained to learn a nonlinear mapping from multidimensional fault-current signatures to the corresponding optimized relay-setting vectors. In the online stage, the trained CNN serves as a predictive surrogate model, rapidly recommending coordinated relay settings for the current operating condition. The framework is validated using a 16-bus distribution system and the IEEE 37-bus test feeder. The results show that the proposed method can restore correct primary–backup relay operating sequences and maintain coordination time intervals (CTIs) within the required 0.2–0.4 s range under the studied DER-rich operating scenarios. This CNN-based preventive calibration approach provides a rapid, adaptive decision-support tool to improve protection coordination robustness against DER-induced operating uncertainties. Full article

This study investigates the effects of introducing a manufacturer-operated online recycling channel (hereafter “online channel”) and adopting a residual-capacity-based differentiated-pricing strategy in spent power-battery recycling. We develop a manufacturer-led Stackelberg game involving a manufacturer, a retailer, and consumers. We solve and compare the equilibrium decisions under four modes: the offline-only identical-pricing (SI) mode, the dual-channel identical-pricing (DI) mode, the offline-only differentiated-pricing (SH) mode, and the dual-channel differentiated-pricing (DH) mode. The results show that under the baseline model, the total collection volume under the dual-channel recycling mode is greater than that under the offline-only recycling mode, while the retailer’s offline collection volume is lower. This indicates that introducing an online channel expands total collection but diverts part of the retailer’s offline collection. Compared with uniform pricing under the same channel structure, differentiated pricing yields the same total collection volume but performs better in collecting batteries with high residual capacity, suggesting that it reallocates incentives toward high-residual-capacity batteries rather than increasing total collection. For mode preference, the manufacturer earns the highest profit under the DH mode, whereas the retailer achieves the highest profit under the SH mode. These results are further examined in several extended studies. The contributions are threefold. First, the study distinguishes the market-expansion effect of online recycling from the value-recognition effect of differentiated pricing. Second, the consistency and applicability boundaries of baseline conclusions are examined and proposed under operational costs, residual-capacity measurement errors, consumer differentiation, and piecewise-linear demand. Third, a feasible profit-sharing mechanism is developed to coordinate the conflict between the manufacturer’s and retailer’s mode preferences. Full article

Accurately modeling learners’ course preferences is a key task in Massive Open Online Course (MOOC) recommendation. Existing sequential recommendation methods can capture recent changes in learners’ interests but have limited ability to exploit shared preference patterns across learners. In contrast, graph-based recommendation methods can capture collaborative information from learner–course interactions, but they often ignore the effects of behavior order and time intervals on interest evolution. To address these problems, this study proposes Sequence Graph Fusion Recommendation (SGFRec), a method for MOOC course recommendation. SGFRec models learners’ recent interests from course-taking sequences and collaborative preferences from learner–course interactions. It further considers the time intervals between course interactions to distinguish different learning rhythms. An adaptive fusion mechanism then balances recent individual interests and collaborative information for each learner. Experimental results on the MOOCCube and MOOCCourse datasets show that SGFRec outperforms several baselines, demonstrating improved offline recommendation ranking performance. Full article

While e-commerce has transformed consumption patterns, online shopping behavior remains influenced by the physical environment. Using Shanghai as a case study, this research applies machine learning, SHAP analysis, and K-Means clustering to examine the nonlinear impacts and spatial variations in the built environment on e-shopping. The findings reveal that: (1) E-shopping expenditure follows a long-tail distribution and displays a concentric spatial pattern, peaking between the Outer and Suburban Rings while decreasing within the Inner Ring and beyond the Suburban Ring. (2) Built-environment factors exhibit non-linear effects, with local shopping potential and transit distance playing dominant roles. Indicators such as store density and delivery facility coverage show inverted U-shaped threshold effects, indicating a shift from complementarity to substitution between offline and online retail. (3) The urban space can be clustered into three sub-district types—traditional residential, single-function, and mixed-use—each with distinct e-shopping patterns and drivers. This research highlights the spatial mechanisms shaping digital consumption, providing empirical evidence for context-specific retail planning in megacities. Full article

Up/down counters are ubiquitous in the alarm and fault-confirmation logic of electro-mechanical systems. In aircraft, several electro-mechanical modules provide position feedback for flight control; the Linear Variable Differential Transformer (LVDT) is a representative one, converting mechanical displacement into an electrical signal whose reliable monitoring is critical to flight safety. As such counters are deployed in ever more complex systems and more uncertain environments, rising safety requirements render their heuristic tuning unreliable. To address this challenge, this paper proposes a quantitative, reliability-oriented procedure for counter-based monitors, which replaces heuristic parameter tuning. Both healthy and faulty signal distributions are estimated by Kernel Density Estimation (KDE), so the framework handles non-Gaussian noise and FMEA-weighted failure modes. The threshold-and-counter logic is modeled as a finite-horizon absorbing Discrete-Time Markov Chain (DTMC), which yields the false-confirmation probability, missed-detection probability, and detection delay over a bounded horizon instead of long-run rates. Thresholds and counter parameters are then synthesized offline, leaving a lightweight online monitor that needs only threshold comparison and integer counter updates. We evaluate the method on an LVDT sum-voltage monitor using real aircraft healthy measurements and Simulink-based fault injection with 45 detectable modes, assessing the synthesized monitor on $\mathrm{558,001}$ real measured samples and an FMEA-driven fault population. Results show that, among the compared confirmation logics, our workflow yields a counter that meets the ${10}^{-9}$ false-confirmation target and the ${10}^{-6}$ missed-detection target while attaining the lowest detection delay. Full article