Search Results (11,548)

Electrocatalytic seawater splitting is an ideal strategy for the large-scale production of green hydrogen. Compared to scarce freshwater, oceanic seawater electrolysis represents a game-changer for the hydrogen economy. Herein, we report a cost-effective one-step synthesis of binder-free, self-supported 3D nickel–manganese–cobalt (NiMnCo) coatings on titanium (Ti) substrates and evaluated their electrocatalytic performance for the hydrogen evolution reactions (HERs) in alkaline media (1.0 M KOH), simulated seawater (SSW, 1.0 M KOH + 0.5 M NaCl) and alkaline natural seawater (ASW, 1.0 M KOH + natural seawater). These ternary coatings were electrodeposited on Ti substrates using an electrochemical deposition method via a dynamic hydrogen bubble template (DHBT) technique. The optimized ternary NiMnCo/Ti-2 electrocatalyst exhibited an enhanced HER activity in both alkaline and seawater media, achieving an ultra-low overpotential of 29, 59 and 66 mV to reach the benchmark current density of 10 mA cm⁻² in SSW, ASW and 1.0 M KOH, respectively. This efficient 3D ternary NiMnCo/Ti-2 electrocatalyst demonstrated stable long-term performance at a constant potential of −0.23 V (vs. RHE) and a constant current density of 10 mA cm⁻² for 50 h without any significant degradation. Furthermore, it exhibited long-term stability in alkaline electrolyte and simulated seawater during multi-step chronopotentiometric testing at variable current densities from 20 mA cm⁻² to 100 mA cm⁻² for 18 h. This superior performance can be attributed to its unique intermetallic structure and multi-component composition, which provides good Cl⁻ resistance, electrochemical stability and synergistic effects among its constituents. Therefore, the optimized NiMnCo/Ti-2 electrocatalyst is a promising candidate for practical seawater electrolysis aiming at green hydrogen production. 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

Granular modeling has emerged as an interpretable framework for nonlinear system representation by constructing clusters of meaningful data units within the input and output domains. Unlike conventional neuro-fuzzy models that yield crisp outputs, granular models generate fuzzy-set-based outputs, preserving uncertainty information. However, traditional granular architectures rely on linear aggregation mechanisms, limiting their expressive power and structural adaptability. This paper proposes a novel framework termed Logic-Based Fuzzy Granular Networks (LFGNs), in which conventional granular models are enhanced through the incorporation of fuzzy logical neurons implementing AND–OR operations. The proposed logic-based structure enables nonlinear interactions among induced granules while maintaining interpretability. To further improve predictive performance, LFGNs are embedded into a boosting framework, forming a boosted LFGN in which each LFGN acts as a weak learner. Extensive simulation studies on benchmark datasets indicate that the proposed approach outperforms conventional granular models and the existing boosting method in terms of regression accuracy. The integration of logical neurons, boosting, and fuzzy granular models provides a unified and robust granular modeling framework. Full article

The increasing demand for innovative strategies in sustainable food production—driven by rapid global population growth, particularly in sub-Saharan Africa (SSA)—necessitates urgent attention to agricultural resilience. Recent technological advancements have enhanced crop productivity, post-harvest preservation, and environmentally sustainable farming practices. However, three critical bottlenecks remain: (i) the lack of accurate, maize-specific yield prediction methods tailored to SSA; (ii) limited multimodal modeling approaches capable of capturing complex, nonlinear interactions among heterogeneous data sources; and (iii) a lack of explainability mechanisms, which render high-performing models “black boxes” and hinder stakeholder trust. To address these gaps, this study presents an explainable machine learning framework for smart maize yield prediction. We integrate multimodal SSA-specific soil, crop, and weather data to capture the multi-dimensional drivers of maize productivity. Six diverse algorithms—including extreme gradient boosting (XGBoost), light gradient boosting machine (LGBM), categorical boosting (CatBoost), support vector machine (SVM), random forest (RF), and an artificial neural network (ANN) combined with a k-nearest neighbors (kNN)—were benchmarked to evaluate predictive performance. To ensure robustness against spatial heterogeneity, we employed a Leave-One-Plot-Out (LOPO) cross-validation strategy. Empirical results on unseen test data identify CatBoost as the best-performing model, achieving a coefficient of determination of ($R^2 =~76\%$), demonstrating its ability to capture complex, nonlinear relationships in agricultural data. To enhance transparency and stakeholder trust, we integrated Local Interpretable Model-agnostic Explanations (LIME), providing plot-level insights into the physiological and environmental drivers of maize yield. Together, these contributions establish a scalable and interpretable modeling framework capable of supporting data-driven agricultural decision-making in SSA. Full article

This paper analyzes “prompt engineering” through the economic lens of self-insurance against the risk of errors from noisy AI systems. To formalize this approach, we model an agent under cognitive load, allocating effort between working unassisted and prompting an AI assistant. The theoretical model demonstrates that an agent’s optimal prompting effort is driven by the agent’s attitude toward risk. Specifically, the model proves that risk-averse agents rationally “over-invest” in prompting effort, while risk-seeking agents “under-invest” relative to the risk-neutral benchmark. This outcome stems from the covariance between the marginal utility of performance and the marginal product of prompting. This alignment is positive for risk-averse agents, effectively boosting the AI’s perceived productivity. The novel implication is that prompting effort is an economically meaningful behavior that can be informative about an individual’s underlying attitude toward downside AI risk. These results offer a new perspective for understanding heterogeneity in AI adoption and oversight. They also suggest that, under comparable task conditions and controlling for prompting ability, observed prompting effort may be informative about attitudes toward downside AI risk. The framework therefore provides a risk-management perspective for understanding heterogeneity in AI governance in high-stakes settings such as healthcare and finance. Full article

Injection molding process optimization has predominantly relied on simulation-generated data, which cannot capture machine-specific variability and stochastic process noise inherent in real manufacturing environments. This paper presents an experiment-driven machine learning framework for multi-objective optimization of injection molding process parameters targeting volumetric shrinkage, warpage, cycle time, and part weight. Physical experiments were conducted on an industrial injection molding machine using high-density polyethylene with a face-centered central composite design. Systematic benchmarking of four machine learning algorithms under identical cross-validation protocols identified Gaussian process regression as the best-performing surrogate model for the majority of quality metrics, while warpage prediction remained challenging across all algorithms due to its complex thermo-mechanical origins. Permutation-based feature importance analysis established a clear parameter hierarchy, identifying holding time as the dominant factor governing multiple quality responses. Constrained Bayesian optimization with progressive constraint tightening was employed to identify optimal parameter sets and fundamental process capability boundaries. The resulting parameter configurations were validated against a held-out test set. This work demonstrates that rigorous, data-driven optimization using exclusively experimental data provides a viable and practically achievable alternative to simulation-based approaches, contributing to experiment-centric smart manufacturing in polymer processing. Full article

Long-term energy sustainability in the built environment depends not only on deploying renewables but also on maintaining high energy efficiency that consistently lowers demand and enables more effective use of low-carbon electricity over time. This paper presents an illustrative case study that demonstrates a low-data, EPC/audit-based screening workflow for assessing operational energy, carbon, and water-related indicators in a non-residential building. An explanatory case study is conducted for a mixed-use logistics facility in Poland (≈610 m²), combining approaches to useful/final/primary energy indicators with operational carbon and water footprints. The operational water footprint is evaluated as a screening metric (L/kWh) applied to the annual electricity balance and tested across PV self-consumption levels (25/50/75%) to reflect the role of energy management and flexibility. The results indicate that an efficiency-oriented modernization pathway supported by PV integration (≈64 kWp; ~57,350 kWh/yr) reduces the primary energy performance indicator EP from 154 to 62.5 kWh/m²·yr, corresponding to a 59% reduction in annual primary energy demand. The operational water footprint indicator decreases nearly linearly with increasing PV self-consumption, demonstrating that long-term benefits depend on sustained efficiency and on maximizing on-site renewable utilization through controls, demand shifting, and/or storage. Overall, the framework supports transparent benchmarking and the development of staged pathways for integrating renewable and low-carbon energy systems into logistics-building portfolios, while maintaining an analytical focus on operational energy, carbon, and water performances. Full article

This paper presents a hybrid econometric and machine-learning framework for forecasting GDP that bridges long-run structure with short-run regime dynamics. Using annual World Bank data spanning 1960 to 2024, the framework combines three complementary components: an ARIMA baseline that captures persistence, a three-state Hidden Markov Model (HMM) that provides probabilistic regime identification, and an LSTM-based extension that learns nonlinear patterns associated with regime transitions. Detailed out-of-sample forecasting evidence is reported for five representative countries (the United States, China, Germany, India, and Greece), chosen to illustrate performance across different volatility profiles and economic environments. Across these case studies, the integrated HMM–LSTM approach often delivers lower forecast errors than the benchmark alternatives, although the magnitude of the gains is not uniform across countries. Beyond point forecasting performance, the regime layer yields an interpretable probabilistic representation of business cycle conditions that can support real-time monitoring and early-warning assessment. By combining transparency with adaptability, the proposed framework contributes to the forecasting literature and provides a practical decision-support tool under heightened macroeconomic uncertainty. Full article

Urban areas are pivotal to achieving the Sustainable Development Goals (SDGs), yet sustainability monitoring at the municipal level remains fragmented, difficult to operationalize, and weakly comparable across cities. Although the SDGs provide a comprehensive global agenda and ISO 37120 offers a standardized set of city indicators, municipalities still face practical barriers in translating global targets into actionable, jurisdiction-sensitive, and measurable metrics aligned with local responsibilities and available data. This study addresses this gap by presenting the design of an integrated, target-level urban sustainability assessment framework grounded in SMART (Specific, Measurable, Achievable, Relevant, and Time-bound) principles and explicitly tailored to municipalities in developed-country contexts. The framework contributes (i) a structured procedure for disaggregating and reallocating SDG targets according to municipal responsibilities, (ii) a six-dimension architecture that consolidates SDG targets and ISO 37120 themes into a coherent, governance-oriented structure (Government and Economic Development; Civic & Social Infrastructure; Environment and Climate; Infrastructure and Urban Planning; Health; Urban Living Conditions), and (iii) a SMART-based indicator screening logic that prioritizes feasibility, data availability, and benchmarking potential, thus supporting the green transition in Urban Areas. The framework is empirically examined through validation against sustainability reporting practices of the Porto City Council, quantifying indicator coverage, assessing alignment with municipal mandates, and identifying systematic gaps—particularly in cross-cutting areas such as governance transparency, equity monitoring, and long-term climate adaptation. Overall, the results indicate that the proposed approach strengthens coherence, measurability, and comparability in urban sustainability assessment, supporting evidence-based municipal decision-making, performance benchmarking, and more strategically aligned SDG localization. Full article

Satellite–terrestrial integrated networks provide seamless global coverage, especially in remote areas where terrestrial deployment is costly. Integrated sensing and communications (ISAC) enhances spectral efficiency by merging both functions on a single platform. This paper proposes a novel integrated satellite–terrestrial ISAC architecture, where a satellite performs simultaneous communication and sensing. The satellite transmits communication signals and sensing waveforms to an Earth Station, which then relays them to a terrestrial base station to serve multiple users. We formulate a joint beamforming design problem to maximize the sum rate of users under quality-of-service constraints, backhaul capacity limits, beampattern requirements for sensing, and power budgets. With perfect channel state information, the non-convex problem is transformed into a difference-of-convex form and solved via the convex–concave procedure. For imperfect channel state information, a robust method combining successive convex approximation and the S-procedure is developed. Simulations show the proposed design outperforms benchmarks and is suitable for low-Earth orbit satellite systems. Full article

In harsh marine environments, vessel energy storage systems (VESS) face elevated thermal runaway (TR) risk, yet practical early warning remains difficult because early voltage differences between TR high-risk cells and normal cells are often weak, warning thresholds vary across operating segments, and decisions relying on a single feature are prone to false or missed warnings. To overcome these difficulties, this study develops a four-part early warning strategy for TR high-risk cells in VESS. First, the original cell voltages are denoised through multiscale jump plus mode decomposition and Spearman correlation guided mode reconstruction to suppress irrelevant interference. Second, an improved Sigmoid nonlinear mapping is introduced to enhance subtle inter-cell voltage deviations and improve early separability. Third, sparse representation is used to construct a cell deviation score, and an adaptive threshold is employed to perform primary abnormal-cell screening under varying segment conditions. Finally, multidimensional mutual information value derived from voltage, temperature, and their rates of change is incorporated into a joint assessment methodology to further verify the abnormal state of flagged cells. Validation on 18 independent real operation cases comprising 2483 discharge segments shows that, across the evaluated TR high-risk cases, the shortest confirmed warning lead time achieved by the proposed strategy was 14 days. The proposed strategy also reduced false and missed warnings, outperformed the compared benchmark methods overall, and retained computational feasibility for onboard application in VESS. Full article

Analyzing large-scale student feedback is critical for higher education quality assurance, yet manual analysis is inefficient and subjective. This paper proposes an integrated framework that unifies supervised classification, unsupervised clustering, and retrieval-augmented generation (RAG) to produce evidence-grounded and actionable insights. Ensemble-based supervised models perform thematic classification, while multi-encoder embedding fusion enables unsupervised discovery of coherent feedback clusters. A multi-stage RAG module integrates category predictions and cluster structure to retrieve representative evidence and generate transparent summaries with citation traceability. The framework is evaluated on student feedback collected from a Central Asian university and two public benchmarks, EduRABSA and Coursera course reviews, covering seven thematic categories. The supervised ensemble achieves 83.0% accuracy and 0.829 Macro-F1 on the primary dataset, while unsupervised clustering attains a silhouette score of 0.271 under the best fusion strategy. Independent evaluation on external benchmarks yields ensemble accuracy of 81.1% on EduRABSA and 49.8% on Coursera, confirming the framework’s adaptability across diverse educational contexts. By leveraging supervised labels and unsupervised structure, the proposed framework enables evidence-grounded, category-aware LLM-based summaries that faithfully reflect the diversity and distribution of student feedback and support actionable educational decision-making. Full article

EEG-based classification of mental disorders has increasingly relied on deep learning models, which are computationally intensive and difficult to interpret, limiting reproducibility and clinical deployment in resource-constrained or cross-site settings. We propose a configurable and modular machine learning framework for EEG-based classification that emphasizes interpretability, flexibility, and rigorous evaluation using schizophrenia detection as a representative use case. Our framework integrates standardized preprocessing, multiscale feature extraction, minimum redundancy–maximum relevance feature selection, and configurable ensemble learning. It also supports multiple validation strategies, including random splits, k-fold cross-validation, and leave-one-subject-out (LOSO), enabling systematic assessment of subject-level generalization. We evaluated the framework on two open EEG datasets: Warsaw IPN (Institute of Psychiatry and Neurology, 19 channels, 250 Hz; 28 subjects) and a Moscow adolescent cohort (16 channels, 128 Hz; 84 subjects). Results show that validation strategy strongly affects model performance. While K-fold validation yielded epoch-level accuracies of 98.06% and 91.47%, LOSO results were much lower: 76.12% (epoch-level) and 82.14% (subject-level) for Dataset 1, and 70.71% (epoch-level) and 77.38% (subject-level) for Dataset 2. These findings demonstrate the risk of overestimated performance due to data leakage and underscore the importance of subject-independent evaluation. Our proposed framework provides a low-complexity, interpretable, and extensible benchmark for reproducible EEG-based machine learning, with interpretable feature representations linked to EEG dynamics and potential applicability to broader neuroengineering and clinical decision-support systems. Full article

Energy management of multi-carrier energy hubs (MCEHs) is a challenging task, particularly when fuel cell electric vehicle (FCEV) stations are included, due to the stochastic nature of FCEV demand, system loads, and integrated renewable energy resources (RERs) such as wind turbines (WTs) and photovoltaic (PV) systems. This paper aims to optimize the energy management of an MCEH-based microgrid to simultaneously minimize total operating costs and emissions. To this end, a novel enhanced quadratic interpolation optimization (EQIO) algorithm is proposed. The proposed EQIO algorithm incorporates two key improvements: a best-to-mean quasi-oppositional-based learning (BMQOBL) strategy and an evaluation mutation (EM) strategy. The performance of EQIO is evaluated using the CEC 2022 benchmark functions, and the obtained results are compared with those of other optimization techniques. Three case studies are investigated: (i) energy management of the MCEH microgrid without RERs, (ii) sustainable operation (with RERs), and (iii) sustainable operation with RERs combined with the application of demand-side response (DSR). Moreover, the proposed framework explicitly supports long-term sustainability goals by enhancing renewable energy utilization, reducing the carbon footprint, and promoting cleaner transportation through efficient integration of FCEV infrastructure. The results demonstrate that integrating RERs reduces operating costs and emissions by 51.47% and 59.69%, respectively, compared to the case without RERs. Furthermore, the combined application of RERs and DSR achieves cost and emission reductions of 55.26% and 53.93%, respectively, compared to the case without RERs. Full article

Person re-identification (ReID) under diverse weather conditions remains a critical yet insufficiently explored problem. Most existing ReID approaches are developed and benchmarked on clear-weather datasets, resulting in significant performance degradation when deployed in rainy, snowy, or hazy environments. Conventional image restoration methods, typically optimized for low-level image quality metrics, are often misaligned with the objectives of high-level identity discrimination and thus fail to improve the person ReID performance. To address these limitations, we propose DW-ReID, a unified framework that integrates weather-degraded image restoration with person re-identification tasks. The proposed DW-ReID is built upon a large-scale Contrastive Language-Image Pre-training (CLIP) model and achieved by a two-stage training paradigm. In the first stage, a set of learnable text prompts is optimized to construct identity-specific ambiguous descriptions for each person’s identity. In the second stage, the optimized text descriptions, together with a frozen text encoder, provide language supervision to jointly train a weather encoder, an image restorer, and a ReID encoder in an end-to-end manner. The experimental results on two our contributed synthetic datasets consistently demonstrate the effectiveness and superior performance of the proposed DW-ReID method. Full article