Search Results (221)

Computing capacity has become a metered factor of production for digitally intensive enterprises, yet its consumption exhibits strong temporal heterogeneity—tidal intraday cycles, weekly contrasts, seasonal surges, and project-driven regime shifts—so that individually provisioned capacity is structurally underutilized. This paper proposes a park-level computing power sharing center (CPSC) as an institutional mechanism that converts the temporal complementarity of co-located enterprises into measurable cost savings. We develop a general mode-design framework that separates CPU core-hours from GPU card-hours, characterizes demand via deterministic tides and stochastic modulations, and derives optimal pooled capacity commitments through a newsvendor-type quantile condition. A parametric calibration protocol maps observable temporal features—peak-to-trough ratios, inter-tenant phase spreads, and residual volatility—into closed-form diversity-factor expressions with Monte Carlo confidence intervals. The procurement model covers a multi-option contract menu (on-demand, one–three-year reserved instances, savings plans, and spot), region-specific pricing, hardware class tariffs, and ancillary costs, including network egress, migration, and data sovereignty compliance; benefits are measured relative to each tenant’s individually optimal reserved portfolio, not naive retail procurement. A mechanism design analysis incorporating Shapley value allocation, Bayesian incentive compatibility, and penalty structures ensures individual rationality and robustness to misreporting and strategic load shifting. We further develop an energy model—with utilization-dependent power draw, facility PUE, embodied carbon, and marginal grid emission factors—showing that financial savings translate into genuine emission reductions only when pooling enables physical capacity retirement rather than mere billing reallocation. The framework is applied to twenty representative Chinese parks spanning seven functional categories; all park-level data are reconstructed from public sources using the calibration methodology, and the reported figures are model-derived projections, not empirical measurements. The model yields procurement saving estimates of 4.6–20.2% relative to individually optimal reserved-procurement portfolios, with high-diversity parks at the upper end. Sensitivity analyses across regional tariffs, hardware mixes, and cross-country utilization benchmarks (Uptime Institute, US DOE, EU Commission) confirm robustness and delineate boundary conditions. This paper concludes with a data provenance taxonomy and a phased implementation roadmap. Full article

While Artificial Intelligence (AI) has transformed materials discovery, the primary bottleneck to technological impact remains the transition from lab-scale synthesis to robust, industrial-scale manufacturing. Most promising materials perish in this depth, which is referred as the “valley of death”. The current review consolidates and critically evaluates the emerging ecosystem of AI-driven strategies and frameworks designed specifically to bridge this “lab-to-fab” gap. The review shifts our attention from property prediction to the engineering-driven problems of manufacturability. Furthermore, the review discusses the main obstacles to scaling the production of materials, such as reproducibility, process optimization in the context of uncertainty, and techno-economic viability, as well as the AI approaches being developed to overcome them. This includes Natural Language Processing (NLP) for method extraction, graph neural networks for reaction modeling, reinforcement learning for process control, Bayesian optimization for definition of process windows, and integrated AI–Techno-Economic Analysis (TEA) frameworks. Equally importantly, the review examines the principal failure modes that constrain practical deployment, including out-of-distribution generalization, incomplete and non-transferable literature-derived data, simulator-to-plant mismatch, uncertainty miscalibration, and the continued need for expert oversight. The article concludes with a forward-looking roadmap for the future of AI in chemical and materials engineering. The proposed conclusion is defined by a paradigm shift from simply finding new materials to creating viable, economical, and scalable pathways to produce them, thereby enabling a new era of synthesis-aware materials innovation. Full article

Solar thermal energy storage systems play a vital role in bridging the gap between intermittent solar availability and continuous energy demand; however, their efficiency is severely constrained by the inherently low thermal conductivity of phase change materials. Integrating physical heat transfer enhancements, such as radial fins, offers a practical solution, but evaluating these non-linear thermal dynamics across diverse design configurations typically incurs heavy computational costs. To address this challenge, this research investigates an artificial intelligence-based predictive framework capable of accurately modeling complex phase change dynamics in fin-configured storage tanks. Utilizing high-fidelity 2D Computational Fluid Dynamics simulation data of a stainless-steel double-tube storage tank filled with RT-50 paraffin wax across 10, 20, and 29 fin configurations, a Multi-Layer Perceptron Artificial Neural Network trained with the Bayesian Regularization algorithm was developed. The model predicts liquid fraction, latent heat distribution, and buoyancy-driven natural convection (Reynolds number) based on fin count and time. The optimal architecture, featuring 30 hidden neurons, achieved exceptional predictive precision, yielding a coefficient of determination of 0.99999, along with individual Mean Squared Error values of 1.29 × 10⁻³ for liquid fraction, 5.73 × 10⁻¹ for latent heat distribution, and 2.64 × 10⁻⁵ for Reynolds number, with average prediction deviation rates consistently below 0.5%. These results demonstrate that high-precision surrogate modeling can effectively replace computationally intensive numerical simulations, offering significant practical implications for the real-time thermal monitoring, rapid design optimization, and intelligent control of advanced solar energy storage technologies. Full article

The safe and stable operation of coal-fired generating units is of critical strategic importance for ensuring the reliable supply of power systems. However, the fault evolution of industrial thermal systems exhibits the characteristics of strong nonlinearity and a long incubation period, coupled with the extreme scarcity of key fault samples (Few-shot) in actual production, which severely limits the engineering application of traditional data-driven diagnostic methods. Existing deep learning models, which are highly dependent on massive and balanced labeled data, not only struggle to overcome the overfitting bottleneck in scenarios with scarce fault samples, but also frequently introduce severe label noise (Label Noise) by ignoring the physical incubation period of faults, resulting in the degradation of the model’s decision boundary. To address the above challenges, this paper proposes a novel fault diagnosis framework integrating multi-scale spatiotemporal feature engineering and the Tabular Prior-Data Fitted Network (TabPFN). Starting from the physical mechanism of the system, this paper develops a dynamic label cleaning strategy based on multivariate statistical deviation, which accurately defines the fault divergence point to eliminate the noise in the incubation period. The constructed multi-scale spatiotemporal feature engineering integrating first-order difference and sliding window statistics can effectively map the transient mutation and steady-state evolution trend of the system. The introduced pre-trained TabPFN model based on the Transformer architecture, relying on its Bayesian inference capability and in-context learning (In-Context Learning) mechanism, can realize parameter-tuning-free and efficient classification for scarce samples. Experiments based on high-fidelity dynamic simulation data from GE Steam Power show that under the strict setting of limiting the training set to only 2000 samples, the proposed method achieves a comprehensive diagnostic accuracy of up to 99.29% and an F1-score of 0.9929 for seven typical operating conditions. Multi-dimensional comparative experiments and ablation studies confirm that the proposed framework comprehensively outperforms six mainstream baseline models, including XGBoost and SVM, in terms of precision, recall, and anti-interference robustness, and also delivers outstanding performance when benchmarked against deep learning models. This provides a brand-new theoretical perspective and technical paradigm for equipment health management in the context of industrial big data. Full article

The large-scale integration of renewable energy has compelled coal-fired power plants to operate under deep peak-shaving conditions, significantly increasing the failure risk of Circulating Fluidized Bed (CFB) boiler heating surfaces due to severe thermal and pressure fluctuations. To address the limitations of traditional static risk evaluations, this study proposes a novel dynamic risk assessment framework integrating a Spatial–Temporal Graph Convolutional Network (STGCN) and a Dynamic Bayesian Network (DBN). The STGCN, enhanced with an operation-adaptive dynamic cross-attention delay module, predicts spatiotemporal temperature and pressure variations across the high-temperature heating surfaces. The predicted variables are incorporated into the DBN as dynamic evidence, which utilizes Noisy-OR logic and an embedded Weibull physical degradation model to continuously quantify cumulative failure probabilities. Case study results demonstrate that the STGCN outperforms traditional LSTM and RNN baselines in prediction accuracy. Furthermore, the DBN effectively maps the distinct degradation characteristics of individual boiler components, accurately identifying the water wall and superheater as having the highest failure risks and the largest fluctuations in marginal failure probability during rapid load cycling. This integrated data-driven approach provides highly accurate, real-time risk predictions, offering essential decision-making support for the predictive maintenance and safe flexible operation of CFB boilers. Full article

Optimizing particle grading for dense granular packing remains challenging because the formulation space expands rapidly with increasing particle-size complexity, while direct evaluation using the discrete element method (DEM) is computationally expensive. In this work, a data-driven framework integrating DEM, Gaussian process regression, and Bayesian optimization is developed for efficient screening of ternary particle-grading formulations under a common high-density condition. Normalized residual interparticle overlap is introduced as a fixed-density measure of geometric incompatibility, which enables different formulations to be compared without repeatedly determining their maximum packing states. The surrogate-assisted search increases the occurrence of low-overlap candidates and shows that favorable formulations are distributed among several regions of the grading space rather than converging to a unique optimum. Repeat evaluations further identify two low-overlap strategies with distinct particle-size distributions and stable packing responses. Contact network and particle rearrangement analyses show that low overlap can be achieved through either a coarse–intermediate framework assisted by local fine-particle accommodation or a more distributed multiscale contact network. Fine particles provide most of the configurational mobility during relaxation, while the distribution of this mobility depends strongly on the underlying grading architecture. These findings establish practical relationships among particle-size hierarchy, distribution breadth, component fraction, and packing response, which can guide the selection of dense powder formulations and reduce reliance on extensive trial-and-error screening. Full article

Unmanned aerial vehicle (UAV) accidents pose growing challenges to public safety and airspace management. This study develops an open-data-driven hierarchical Bayesian network (BN) with a Leaky Noisy-OR mechanism to analyse factors associated with consequence severity among recorded UAV accidents, and extends it to a Bayesian decision network for response simulation. Using 633 public accident records and matched meteorological data, 14 binary risk-factor nodes spanning human, machine, environmental, and management dimensions were constructed. Stratified five-fold cross-validation yielded a mean validation F1 score of 0.922 and an AUC of 0.784. Backward inference ranked airspace exposure, wind, and operation error highest under severe-consequence conditioning, whereas sensitivity analysis identified wind, bad weather history, and operation error as the most influential root-node parameters. Under the assumed directed acyclic graph (DAG), the bad weather history→weather→environment→risk state path had the highest average edge-influence score (0.853). Under the baseline safety-priority assumptions, the reroute strategy was preferred, yielding the highest expected utility (31.967) and reducing the model-estimated post-decision high-risk probability from 81% to 45%. Alternative preference settings ranked the adjust strategy first. The framework integrates open-data severity analysis with assumption-explicit response simulation. Full article

The quality and safety supervision of agricultural products at the grassroots level has long faced the triple superposition dilemma of small-sample sampling, multi-source evidence conflict, and risk chain evolution. Although existing data-driven models have considerable accuracy, they are difficult to leverage for intervention decisions, and the simple serial connection of traditional Bayesian networks and evidence theory cannot respond to dynamic scenarios. Aiming at this research gap, this paper constructs a dynamic risk assessment model, CIBE-DR, that deeply couples Bayesian networks with evidential reasoning. It contains three core innovations. First, the structure learning method of the causally identifiable Bayesian network embeds a graded do-calculus identifiability score covering both back-door and front-door criteria into the BDeu scoring function and combines this reward with an expert-prior divergence penalty that breaks Markov equivalence so as to realize the transition from relevance modeling to intervention decision modeling. Second, the conflict-aware adaptive evidence synthesis rule orthogonally decomposes multi-source conflict into an epistemic component and an ontological component, which are modeled respectively by Tsallis belief entropy and abductive inference over a discrete twenty-seven-point heterogeneity hypothesis space and are then fused under a reparameterized Dempster–Yager interpolation in which the two endpoints recover the two named rules under a single consistent interpretation. Third, the bidirectional closed-loop coupling mechanism between BN and ER realizes the mutual calibration between the conditional probability table and the evidence credibility prior under a Lyapunov monotone descent argument with the explicit Lipschitz bound $L_{\theta }$ ≤ 0.028 < 1, endowing the model with time-varying self-correction ability. Based on experiments on 156,847 sampling samples from counties and townships in East China, Central China, and Southwest China from 2021 to 2024, the proposed method achieved the best value in six of the seven evaluation indicators, with a minority recall of 0.864 ± 0.014, an intervention effect estimation error of 0.063 ± 0.005, and a dynamic response delay of 2.8 ± 0.3 days, significantly ahead of eleven mainstream baselines under the McNemar test on classification ($p$ < 0.001) and the Wilcoxon signed-rank test on intervention-effect estimation ($p$ < 0.001). The only indicator on which CIBE-DR does not lead is overall accuracy, which is 0.002 lower than that of Transformer; this difference does not reach statistical significance under the McNemar test ($p$ = 0.32) and does not weaken the value of grassroots supervision in the strong-imbalance scenario where the positive rate is only 1.04%. The robustness advantage of the model is particularly prominent in the scenarios of sparse data, adversarial perturbation, and prior-graph incompleteness, and the intervention-effect estimates were additionally validated against two post-2022 policy interventions with absolute deviations of 1.4 and 1.2 percentage points respectively. These results verify the product gain and grassroots deployability of the three mechanisms. Full article

Reliable navigation in GPS-denied environments remains a critical challenge for autonomous vehicles (AVs), particularly in complex indoor and urban settings. GPS-based localization systems often fail under these conditions, highlighting the need for resilient multimodal solutions. In this article, we present a radar-assisted tracking system that integrates LiDAR and inertial measurements within a sensor-fusion architecture to achieve robust navigation. The principal methodological contribution is a unified tracking and prediction framework that combines Bayesian state estimation with learning-based temporal prediction, enabling accurate tracking while continuously forecasting the slave robot’s short-term future state from mapping observations generated by the master robot, with a typical end-to-end perception-to-action latency of 20–60 ms. The communication and prediction forecasting module operates with an update interval below 35 ms, enabling real-time cooperative robotic operation. Sensor data are fused through a pipeline incorporating Gaussian Mixture Models (GMMs) for post-processing, which helps mitigate the limitations associated with individual sensors during edge processing. Moreover, Kalman filtering is employed to mitigate sensor noise and drift, thereby improving state estimation accuracy through trajectory smoothing. The fused spatiotemporal information is subsequently exploited by a Convolutional Recurrent Neural Network (CRNN) coupled with a Nonlinear Autoregressive model with eXogenous Inputs (NARX) to model the robot’s motion dynamics and provide short-horizon state prediction. Through simulations and real-world indoor experiments conducted in GPS-denied environments, we validate the system’s ability to provide accurate and continuous pose estimation with low localization errors. Experimental results show that the proposed framework achieves root-mean-square errors of 0.12 m, 0.15 m, and 0.28 m along the X, Y, and Z axes, respectively, while maintaining sub-meter maximum position deviations throughout the evaluated trajectories. These results confirm that the proposed framework provides reliable localization and predictive state estimation for cooperative robotic navigation in indoor GPS-denied environments. Future work will investigate outdoor validation and extend the framework to additional data-driven decision-making models for future robotic services. Full article

Artificial intelligence (AI) is reshaping the architecture, engineering and construction (AEC) sector. However, AEC education remains rooted within traditional disciplinary boundaries and a technology-centric training model, creating a widening mismatch between graduates’ capabilities and the cognitive, collaborative, and interdisciplinary demands of AI-enabled practice. Situated within the Chinese higher education context, this study bridges this gap through a multi-stakeholder survey (n = 352) noting perspectives from academia, industry and research. One-way ANOVA and Tukey’s HSD test were used to examine differences across stakeholder groups and disciplines, while a Bayesian Network was developed to model competency pathways, simulate intervention scenarios and identify key leverage points for curriculum reform. The analysis reveals that meaningful AI integration requires a reconstruction of competency, rather than the mere addition of standalone technical or software courses; it calls for fundamental changes in professional formation, curricula and pedagogy. Four core competencies emerged from the data: professional expertise, systems thinking, interdisciplinary collaboration and AI-enabled problem-solving. Bayesian Network simulations further indicated that curriculum expansion alone improved AI knowledge acquisition by 39.6%, but yielded only modest gains in practical skills (13.9%) and application competencies (4.5%). By contrast, integrated interventions that combined teacher development, university–industry collaboration and project-based practice produced substantial improvements in AI application competencies (37.9%), employment adaptability (29.3%) and industry satisfaction (14.1%). These divergent findings highlight the necessity of coordinated educational interventions to reconcile stakeholder expectations and foster AI-oriented competency development. Based on this evidence, the study proposes a competency-oriented framework and a phased curriculum transformation pathway, providing an empirical foundation for AI-driven curriculum reform and competency reconstruction in AEC education. Full article

Dissolved oxygen (DO) dynamics in river systems exhibit complex nonlinear and non-stationary characteristics driven by interactions among meteorological and water-quality factors, posing challenges for accurate prediction and identification of dominant drivers. To address these challenges, this study establishes a VMD-Bayes-LSTM model by integrating Variational Mode Decomposition (VMD), Bayesian optimization, and Long Short-Term Memory (LSTM) networks. The datasets used in this study were collected from monitoring data at eight sites in the Yangtze River Basin, including water-quality and meteorological factors. Compared with several benchmark models, the VMD-Bayes-LSTM model achieves the best performance in daily DO concentration prediction, with mean R², KGE, MSE, and MAE values of 0.876, 0.900, 0.208, and 0.247, respectively. Meanwhile, this model supports multi-step prediction and achieves five-day-ahead forecasting of DO concentrations. Based on this, a SHAP value-interpretable VMD-Bayes-LSTM model is constructed. Analysis indicates that water temperature and average air temperature are the primary predictive factors at the CC, PT, DQ, LS, NJG, and ZT sites with a cumulative contribution rate exceeding 60%, reaching a maximum of 90% at the PT site; notably, at the LD and LJG sites, turbidity emerges as the dominant water-quality parameter, accounting for 59.6% and 36.2% of the predictive contribution, respectively. Full article

Quick Access Recorder (QAR) data provide high-dimensional onboard flight records that are increasingly used for data-driven aviation safety analysis. As flight operations become more complex, conventional manual monitoring and threshold-based exceedance detection are often insufficient for identifying evolving risks in a timely and interpretable manner. This paper reviews QAR data-driven methods for flight anomaly detection and risk warning from the perspective of statistical learning and aviation data science. First, the main characteristics of QAR data are summarized, including multi-source heterogeneity, temporal dependence, missing values, noise, and severe class imbalance. Common preprocessing techniques, such as missing-value imputation, trajectory correction, feature engineering, downsampling, and imbalanced-data handling, are then reviewed. Second, existing methods are organized into four groups: statistical monitoring and rule-based methods, clustering and unsupervised anomaly detection, Bayesian and probabilistic risk reasoning, and hybrid machine learning with explainable AI. Representative approaches include statistical process control, association rules, Gaussian mixture models, CurveCluster, Fast-DTW, Bayesian networks, dynamic Bayesian networks, VAE-LSTM, MAD-XFP, and XGBoost with SHAP interpretation. The review further discusses typical applications in landing risk warning, takeoff risk assessment, flight operation pattern recognition, and aviation noise prediction. Finally, key challenges are summarized, including model interpretability, real-time deployment, cross-aircraft and cross-airport generalization, data quality, causal reasoning, and privacy-preserving collaboration. This review provides a structured reference for using QAR data to support aviation safety assessment, risk warning, and operational decision-making. Full article

Machine learning has created new opportunities for gas adsorption research using nanoporous materials, but the field’s evolution remains insufficiently quantified. This study retrieved literature from the Web of Science Core Collection for 2010–2026 and retained 730 valid records from 1581 initial publications after screening. VOSviewer, CiteSpace, and R were used to analyze publication growth, collaboration networks, journal sources, and thematic evolution. Results show that annual output remained generally below 20 before 2019, then increased rapidly and reached approximately 280 publications in 2025, indicating accelerated integration of machine learning with adsorption simulation, material screening, and performance evaluation. The source distribution broadened from a limited set of chemistry and engineering journals to diverse venues, with recent high publication weights in Chemical Engineering Journal, Separation and Purification Technology, ACS Applied Materials & Interfaces, Microporous and Mesoporous Materials, and Journal of Materials Chemistry A. Collaboration analysis identified 10 compact author clusters, including groups associated with Randall Q. Snurr, Seda Keskin, Zhiwei Qiao, Qingyuan Yang, and Chongli Zhong, whereas the weak bridging links among clusters indicate that cross-community collaboration remains limited. Country and institutional analyses show that China, the United States, Canada, Iran, India, South Korea, and the United Kingdom are leading contributors, with Guangzhou University, Koç University, Northwestern University, the Chinese Academy of Sciences, Beijing University of Chemical Technology, and the United States Department of Energy occupying prominent positions. Keyword evolution reveals a shift from adsorption behavior and porous adsorbents toward data-guided material selection, high-throughput screening, deep learning, Bayesian optimization, and performance optimization, offering guidance for data-driven adsorbent discovery. Full article

Accurate prediction of creep rupture life is essential for ensuring the long-term reliability of high-temperature components in power generation and petrochemical industries. Selecting appropriate data-driven models for limited and heterogeneous creep datasets remains a critical challenge, as conventional accuracy-based comparisons do not fully capture model behavior under varying service conditions. This study presents a unified evaluation framework for systematically comparing multiple machine learning and deep learning models for creep rupture life prediction of 2.25Cr–1Mo steel. The framework integrates predictive accuracy, prediction reliability, regime-specific error analysis, and computational efficiency, enabling a comprehensive assessment beyond global error metrics. The input feature space is reduced from seventeen to eight physically meaningful variables without loss of predictive performance. To further assess model robustness, prediction errors are analyzed across four distinct rupture life regimes, revealing significant variations in model behavior that are not reflected in aggregate metrics. Results indicate that support vector regression (SVR) provides the most consistent overall performance across all regimes and offers a strong balance between accuracy and computational efficiency. Among deep learning models, a Bayesian neural network (BNN) achieves competitive predictive performance while additionally enabling uncertainty estimation. These findings demonstrate that, for small tabular creep datasets, appropriately regularized models outperform complex neural network architectures, highlighting the importance of matching model complexity to dataset characteristics. This study is limited to a single steel grade, moderate dataset size, and extrapolation beyond trained stress and temperature ranges, which are key directions for future work. Full article

We document a network-level vulnerability in pension-fund systems: lifecycle allocation regulation, designed to protect individual participants, compresses cross-fund return dynamics to the point where provider choice offers little diversification. Using daily net asset value data from second-pillar pension funds in Lithuania over 2019–2025, we found that a single common factor explains more than 72% of total return variance even in the calmest observed periods. We develop an unsupervised regime-detection framework that combines a PCA-based absorption ratio, DTW-based hierarchical clustering, and a Gaussian hidden Markov model with a data-driven crisis threshold. The HMM specification is supported by a dual empirical calibration of the stickiness prior, and its emission estimates agree closely with a fully Bayesian sticky-HMM specification. The framework identifies three latent regimes in which elevated systemic co-movement is the structural norm rather than an exceptional state and shows that funds separate first by lifecycle segment (conservative versus growth cohorts) and only secondarily by provider, with no single label axis reproducing the structure on its own. The absorption ratio has no significant relationship with global equity benchmarks in either calm or High-Concentration regimes, indicating that the detected regimes are not explained by the external benchmarks considered, including global equity indices and a euro-area rate/bond proxy. Cluster-level mean-absolute-active-return amplification of $1.06\times$ to $1.33\times$ during High-Concentration episodes confirms that even conservative funds serving retirement-age participants are not insulated. Full article