Search Results (1,927)

Deep learning models have achieved high accuracy in forecasting problems, but at the cost of large computational energy demand. Brain-inspired spiking neural networks (SNNs) offer a promising, low-power alternative, yet their adoption for time-series forecasting has been limited by information loss from binary spikes and degraded performance in deeper networks. This paper proposes a fully spiking framework that bridges this gap by improving both the encoding and propagation of information in SNNs. The framework introduces a hybrid Delta-Rate encoding mechanism that captures both abrupt changes and gradual trends in time-series data, and a Mem-Spike mechanism that transmits analog membrane potential values to preserve fine-grained information between spiking layers. We further employ residual membrane connections to maintain signal flow in deep spiking networks. Using two public energy load datasets, our enhanced SNNs consistently outperform conventional spiking models, improving prediction accuracy by up to 61.6% and mitigating degradation in multi-layer networks. Notably, it narrows the gap to the selected deep learning baseline (LSTM), achieving comparable accuracy in some settings while requiring only about 10% of the estimated inference energy of that baseline under a common operation-level model. These results show that, within the empirical scope considered here, enhanced conventional SNNs can improve time-series forecasting accuracy while retaining favorable estimated efficiency. Full article

Hybrid supramolecular–nanocomposite hydrogels based on polyethylene glycol (PEG), β-cyclodextrin–adamantane host–guest interactions, and silica nanoparticles represent an important class of hierarchical soft materials with tunable viscoelastic and transport properties. This review critically analyzes recent progress in cyclodextrin–silica hybrid PEG hydrogels, focusing on the mechanistic coupling between stiffness, stress relaxation, and molecular transport arising from the interplay between reversible supramolecular crosslinks and nanoparticle-induced confinement effects. Particular attention is given to how host–guest exchange kinetics regulate dynamic bond rearrangement and affinity-mediated retention of hydrophobic cargo, while silica nanoparticles enhance mechanical reinforcement and modify diffusion pathways through tortuosity and interfacial polymer–particle interactions. The analysis highlights how nanoparticle size, loading level, and surface functionalization influence relaxation spectra and network topology, as well as how environmental stimuli may affect supramolecular bond stability and overall material performance. Comparison with alternative inorganic fillers and mesoporous silica architectures further clarifies the specific advantages of silica in achieving balanced mechanical stability and controlled transport behavior. Overall, current evidence indicates that hybrid CD–silica networks enable partial decoupling of stiffness, relaxation dynamics, and diffusion, although complete independence remains constrained by fundamental polymer physics relationships. These insights support the development of predictive structure–property frameworks for advanced biomedical and controlled release applications. Full article

Accurate demand forecasting and adaptive load balancing are critical for maintaining stability and efficiency in modern smart power grids. This study proposes a hybrid deep learning (DL) framework, termed Transformer-Generative Adversarial Network-Gated Recurrent Unit (Transformer-GAN-GRU), which integrates global attention-based temporal modeling, generative data augmentation, and sequential refinement into a unified architecture. The proposed framework captures both long- and short-term dependencies while improving representation of imbalanced demand patterns. The model is evaluated on three heterogeneous benchmark datasets, namely Pecan Street, the reliability test system-grid modernization laboratory consortium (RTS-GMLC), and the reference energy disaggregation dataset (REDD). Experimental results demonstrate that the proposed model consistently outperforms state-of-the-art baselines, achieving a maximum accuracy (Acc) of 99.49%, a recall of 99.67%, and an area under the curve (AUC) of 99.83%. In addition to high predictive performance, the framework exhibits strong stability, fast convergence, and low inference latency, confirming its suitability for real-time deployment in smart grid environments. Full article

Material considerations are often neglected when developing digital twins, particularly at the relevant length scales that drive material and structural performance. For reinforced composite materials, the microscale has the largest impact on nonlinear material behavior and progressive damage, and thus accurately representing the disordered microstructure of a composite due to processing and manufacturing is critical to developing the material digital twin in the multiscale hierarchy. Automating microstructure characterization is typically done by either training convolutional neural network models using a pretrained encoder or using prompt-based segmentation tools. In this work, a toolset for developing segmentation models is presented, combining these two methods to enable rapid annotation, training, and deployment of microscopy segmentation models for automated material digital twin development without user knowledge of machine learning. Additionally, a Bayesian optimization framework is developed for generating statistically equivalent representative volume elements (SRVE) to a segmented microstructure using a random microstructure generator that implements soft body dynamics. Progressive failure analysis of random, statistically equivalent, and ordered microstructures is compared to the segmented microstructure subject to transverse loading to demonstrate the importance of accurately representing the driving material length scale of a composite digital twin. Ordered microstructures over-predicted crack initiation and ultimate strength and strain. Random and optimized RVE microstructures better agreed with the segmented simulation results, with no significant difference observed between the two methodologies. The improvement in predicted macroscale behavior for models that capture disordered microstructures due to manufacturing processes demonstrates the importance of capturing microstructure features in composites modeling and indicates that SRVEs that capture microstructural features of the physical material can be used in material digital twin development. Further, the toolsets provided in this work allow for rapid development of composite material digital twins without user expertise in machine learning. This has enabled the development of an integrated workflow to automatically characterize and idealize composite microstructures and generate representative geometric models for efficient micromechanics analysis. Full article

Mold infection by Aspergillus and Penicillium spp. in Sithong sweet tamarind (Tamarindus indica L.) during commercial postharvest storage poses quality and food safety risks. However, the current visual detection method, which involves randomly cracking open the pods, is both destructive and laborious. The integration of near-infrared spectroscopy (NIRS) with artificial neural networks (ANN) enables rapid and non-destructive detection while capturing non-linear biochemical–spectral relationships, offering advantages over conventional destructive and linear analytical methods. It was tested as a mold classifier in sweet tamarind pods preserved in commercial ambient conditions (25 °C, 60% relative humidity) for five weeks. Six hundred pods were examined weekly using interactance spectroscopy (800–2500 nm) with six measurement points per pod and four spectral preprocessing methods. The ANN outperformed partial least squares discriminant analysis (PLS-DA) across all storage weeks, peaking at Week 2 with standard normal variate (SNV) preprocessing (prediction accuracy: 85.00%; sensitivity: 0.84; specificity: 0.86; F1-score: 0.85). Advanced tissue degeneration caused spectral heterogeneity, which decreased performance at Week 4 (prediction accuracy: 71.82–76.36%). Principal component loadings identified mold-induced water redistribution and carbohydrate depletion wavelengths at 938, 975–980, and 1035 nm. Week-adaptive calibration is essential for implementation because of the large difference between week-specific model accuracy (up to 85%) and overall storage model accuracy (63.53%). These findings provide a mechanistic underpinning for smaller wavelength-selective sensors and temporally adaptive mold screening systems in commercial tamarind storage. Full article

Modeling the dynamic response of saturated marine soils is crucial yet computationally challenging for traditional methods. Meanwhile, purely data-driven models suffer from sparse data and lack of physical interpretability. To overcome these limitations, this study proposes an intelligent engineering framework based on a frequency-domain physics-informed neural network (FD-PINN) for the forward simulation and inverse parameter identification of saturated seabed soils. Constrained directly by physical laws during the learning process, FD-PINN remains highly reliable even when training data is sparse. By formulating the governing equations in the frequency domain, it directly predicts complex-valued displacement and pore-pressure phasors. Multiscale Fourier feature mappings mitigate spectral bias and capture boundary layers and high-frequency effects. For inverse problems, a phase-sensitive lock-in extraction strategy transforms time-domain measurements into robust frequency-domain targets, enabling the accurate and noise-tolerant identification of poroelastic parameters with clear physical meaning (nondimensional storage parameter S and permeability parameter $\Gamma$). Numerical experiments show that FD-PINN substantially outperforms conventional time-domain PINN, achieving relative $L^2$ errors of ${10}^{-2}\sim {10}^{-3}$ for single- and multi-frequency excitations typical of wave-induced loadings. In particular, $\Gamma$ is consistently recovered with sub-percent relative error, while S can be reliably identified with multi-frequency data. The framework offers a data-efficient, noise-robust approach for high-fidelity modeling and robust parameter inversion, which is particularly valuable in offshore environments where high-quality data is scarce. Full article

Short-term load forecasting plays a pivotal role in modern power system operations yet it remains challenging due to the complex spatiotemporal dependencies in load data. This paper proposes a dual-head attention residual network (DARNet) that significantly advances STLF through three key innovations: (1) a hybrid encoder combining 1D-CNN and GRU architectures to simultaneously capture the local load patterns and long-term temporal dependencies, achieving a 28% better locality awareness than that of conventional approaches; (2) a novel dual-head attention mechanism that dynamically models both the inter-temporal relationships and cross-variable dependencies, reducing the feature engineering requirements; and (3) an autocorrelation-adjusted recursive forecasting framework that cuts the multi-step prediction error accumulation by 33% compared to that with standard seq2seq models. Extensive experiments on real-world datasets from three Chinese cities demonstrate DARNet’s superior performance, outperforming six state-of-the-art benchmarks by 21–35% across all of the evaluation metrics (MAPE, SMAPE, MAE, and RRSE) while maintaining robust generalization across different geographical regions and prediction horizons. Full article

The paper proposed an adaptive preference-based multi-objective optimization framework of intelligent energy management in smart microgrids that are dynamically adapted to operational priorities with regard to real-time grid conditions, stakeholder preferences, and environmental constraints. The suggested hierarchical algorithm combines an improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) with an advanced dynamic preference weight distribution system that can trade off between minimization of operational cost. Reduction of carbon emission, enhancement of voltage stability, enhancement of power quality and maximization of system reliability and adaptability to different operational conditions, such as renewable energy intermittency, demand response schemes and emergencies. The framework presents a new multi-layered preference-learning module that represents the intricate stakeholder priorities in terms of more sophisticated fuzzy logic-based decision matrices, neural network preference prediction, and adaptive reinforcement learning methods and transforms them into dynamic optimization weights with feedback mechanisms. Large-scale simulations on a modified IEEE 33-bus test system coupled with various renewable energy sources, energy storage facilities, electric vehicle charging points, and smart appliances demonstrate superior improvements in performance: 23.7% operational costs reduction, 31.2% carbon emissions reduction, 18.5% system reliability improvement, 15.3% voltage stability increase and 12.8% reduction of deviations in power quality. The proposed system has an adaptive nature with better performance in a variety of operating conditions such as peak demand times, renewable energy intermittency events, grid-connected and islanded operations, emergency load shedding situations, and cyber–physical security risks. The framework is shown to be highly effective under different conditions of uncertainty and variation in parameters and communication delay through intense sensitivity analysis and robustness testing, thus demonstrating its practical applicability in real-world applications of smart grids. Full article

In addressing the issue of insufficient theoretical model accuracy for Cut-out type piezoelectric beams with limiters under the influence of contact-impact nonlinearity, this study utilizes the backpropagation neural network algorithm to develop a data-driven modeling approach based on experimental data from partial distance parameters. This approach aims to achieve accurate predictions of the output voltage and displacement responses of the energy harvester. For different parameter combinations of the limiter gap distance d and installation distance a, amplitude–frequency response data were first systematically collected through experiments, along with time–voltage response data corresponding to different load resistances. Using these data, a training sample set was constructed, and a multi-layer BP neural network prediction model was established with frequency or time as the input and voltage and displacement responses as the outputs. Validation against experimental data demonstrated that the BP neural network can accurately extrapolate and predict the amplitude–frequency response curves of voltage and displacement under various distance parameter combinations, as well as accurately predict the transient voltage outputs under different load conditions. Full article

The capacitance of the DC link is an important variable for the prediction of remaining useful life and failures in frequency inverters. The direct measurement of the DC-link capacitance in inverters operating under load is technically challenging and generally impractical. Recently, a great focus has been given to data-based soft sensors for estimating this variable. These methods, however, are evaluated based only on the estimate errors, and do not take into account the metrological aspects of these estimators. This paper proposes an uncertainty analysis method based on Monte Carlo simulations and bootstrapping that can be applied to all recently published methods for end-of-life (EOL) estimation based on data-driven regression and neural networks. A state-of-the-art model of EOL monitoring based on capacitance estimation was evaluated using the proposed framework, and an experimental study with a frequency converter drive for a brushless DC motor was performed, considering multiple output frequencies, loads and DC-link capacitance conditions. The output distributions are not symmetrical and show that the variable with the most significant impact in the propagated uncertainty is the DC link voltage. The results show confidence interval widths ranging from 12 μF to 61 μF, with wider confidence intervals obtained at higher power setpoints. Full article

Accurate structural analysis is essential for the design and optimization of aircraft wings; however, repeated high-fidelity finite element analysis (FEA) becomes computationally expensive when embedded in iterative design loops. This study presents a machine learning-based surrogate modeling framework for the efficient analysis and optimization of metallic commercial wing structures. A detailed Airbus A320-like wing model was developed and analyzed in ANSYS 2023 R1 under modal, static, and eigenvalue buckling conditions. The general dimensions of the Airbus A320 wing were used only as a reference; the resulting model is a conceptual benchmark rather than a one-to-one geometric replica or a validated digital twin of a specific aircraft wing. Using Latin Hypercube Sampling, 340 high-fidelity samples were generated, with 300 samples used for training and validation and 40 retained as an independent holdout set. The proposed Pyramidal Deep Regression Network (PDRN), a deep learning-based surrogate model whose architecture is automatically tuned using Bayesian Optimization, was benchmarked against Artificial Neural Networks (ANNs), Ensemble Learning, Support Vector Regression (SVR), and Gaussian Process Regression (GPR). On the unseen test set, the PDRN achieved the best overall predictive performance, with RMS errors of 0.8% for mass, 3.1% for the first natural frequency, 11.5% for load factor, and 11.4% for safety factor. To evaluate its practical utility, the trained PDRN was embedded into a PSO-based optimization framework for mass minimization under minimum safety factor, load factor, and first-frequency constraints. The surrogate-guided optimum was verified in ANSYS and remained feasible, yielding a mass of 10,485 kg, a first natural frequency of 1.4142 Hz, a load factor of 1.307, and a safety factor of 1.158. Compared with direct ANSYS in-the-loop optimization, the proposed workflow reached a comparable feasible design with substantially fewer high-fidelity evaluations. These results demonstrate that the PDRN provides an accurate and computationally efficient surrogate for rapid wing analysis and constraint-driven structural optimization. Full article

Grid-forming (GFM) battery energy storage system (BESS) inverters are becoming a cornerstone of resilient microgrids, where severe voltage sags and abrupt operating shifts can challenge both voltage regulation and controller stability. Finite-set model predictive control (FS-MPC) offers fast transient response and multi-objective coordination, yet conventional designs rely on static cost-function weights that are typically tuned offline and may become suboptimal under disturbance-driven regime changes. This paper proposes a forecast-guided KAN-adaptive FS-MPC framework that (i) formulates the inner-loop predictive control in the stationary αβ frame, thereby avoiding PLL dependency and mitigating loss-of-lock risk under extreme sags, and (ii) introduces an Operating Stress Index (OSI) that fuses load forecasts with reserve-margin or percent-operating-reserve signals to quantify grid vulnerability and trigger resilience-oriented control adaptation. A lightweight Kolmogorov–Arnold Network (KAN), parameterized by learnable B-spline edge functions, is embedded as an online weight governor to update key FS-MPC weighting factors in real time, dynamically balancing voltage tracking and switching effort. Experimental validation under high-frequency microgrid scenarios shows that, under a 50% symmetrical voltage sag, the proposed controller reduces the worst-case voltage deviation from 0.45 p.u. to 0.16 p.u. (64.4%) and shortens the recovery time from 35 ms to 8 ms (77.1%) compared with static-weight FS-MPC. In the islanding-like transition case, the proposed method restores the PCC voltage within 18 ms, whereas the static baseline fails to recover within 100 ms. Moreover, the deployed KAN governor requires only 6.2 μs per inference on a 200 MHz DSP, supporting real-time embedded implementation. These results demonstrate that forecast-guided adaptive weighting improves transient resilience and power quality while maintaining DSP-feasible computational complexity. Full article

This study addresses the critical need for sustainable energy by optimizing diaphragm-type piezoelectric elements for efficient waste vibration energy harvesting. Traditional experimental optimization of complex, non-linear design parameters including applied load, tapper diameter, and support structures is often resource-intensive and time-consuming. To overcome these limitations, we developed a novel hybrid machine learning framework that seamlessly integrates an Artificial Neural Network (ANN) with a Modified Grey Wolf Optimization (mGWO) algorithm. The ANN was rigorously trained on experimental data using Bayesian Regularization, establishing itself as a robust and high-fidelity surrogate model capable of accurately predicting voltage output based on diverse input parameters, evidenced by an R-value close to 1. This predictive model subsequently served as the fitness function for the mGWO algorithm, which incorporated a non-linear control parameter to efficiently explore the multi-dimensional design space and effectively balance exploration with exploitation. The framework successfully identified the optimal configuration for maximizing voltage output, predicting a theoretical maximum of approximately 70.67 V. This optimal setup notably involved a high applied load of 100 N, the 6CA multi-pointed tapper configuration, and the three-support boundary condition, which is consistent with the experimentally validated results. The computational findings demonstrated excellent agreement with empirical results while providing significantly higher resolution for design insights. This validated, predictive tool offers a substantial advancement for the future scaling and design optimization of piezoelectric energy harvesters, minimizing the need for extensive physical prototyping and ensuring efficient stress transfer without mechanical failure. Full article

Camber in high-speed railway prestressed concrete (PC) girders increases with service time and affects profile control, ride comfort, and durability; reliable long-term midspan camber prediction is therefore required. Building on established hybrid physics–data modeling and discrepancy-correction ideas, we present a monitoring-oriented two-layer strategy for long-term camber prediction. In the physics layer, a physics-informed neural network (PINN) is formulated in a quasi-static, stage-aware manner to capture the physics-consistent low-frequency trend governed by creep, shrinkage, prestress loss, and staged loading. In the data layer, an XGBoost model learns a bounded, measurement-level residual correction from monitoring features to account for additional effects not explicitly represented in the physics layer, without altering the underlying physics-driven trend. The approach is evaluated using monitoring data from five 1:4 scaled specimens of a 24 m post-tensioned simply supported box girder and is compared against a theoretical calculation and a standalone PINN. Across prediction stages and specimens, the proposed strategy reproduces the measured camber evolution more closely than the benchmarks while preserving physically plausible trend behavior and yielding more consistent errors among girders. These results indicate that, under the present scaled-specimen and independently calibrated setting, a stage-aware physics baseline combined with bounded residual correction can provide closer agreement with the observed camber evolution than the benchmark models under sparse-monitoring conditions. Its engineering applicability can be repeatedly demonstrated across girders with different construction-condition combinations after girder-wise calibration. Full article

Hydropower remains one of the primary sources of power generation. It can be operated as either a base-load or peak-load plant due to its rapid, easy start-up and stop-down capability. However, power plants, old or new, need to be operated and maintained optimally to meet energy demand and maximise economic returns. While the older plants without digital controls such as the Supervisory Control and Data Acquisition (SCADA) system are unable to leverage the evolving technology including big data and Artificial Intelligence (AI), the newer plants or plants that already have some form of data acquisition system have the advantage of leveraging the newer platforms for efficient operation, monitoring and fault diagnosis. Thus, an Artificial Neural Network (ANN), a machine learning (ML) algorithm, was chosen for this case study to predict the generator’s operational stator temperature by selecting six parameters that could potentially affect it. Real data from the 336 MW Chhukha Hydropower Plant (CHP) in Bhutan were used to train the ANN. The prediction of temperature using an ANN in MATLAB^® yielded an R² (correlation coefficient) of 96.8%, which is impressive but can be further improved through various optimisation and tuning methods with increased data volume and complexity. The performance of ANN prediction was validated against other regression models, and the ANN was found to outperform them. This demonstrated its capability to predict and detect generator temperature faults before failures, thereby enhancing hydropower operation and maintenance (O&M) efficiency. The model’s interpretation was also done through Shapley Additive ExPlanations (SHAP). Full article