Search Results (949)

We designed a supervised machine learning framework to predict standard Gibbs free energies, ΔG°, of formal hydrogen atom transfer (f-HAT) for phenolic antioxidants across different radicals and media, enabling rapid and chemically interpretable screening. We curated a DFT dataset of 71 molecules (phenolic compounds and anthocyanidins), with 207 reaction sites, 10 radical reactive oxygen/sulfur species, and three environments (leading to a total of 6210 ΔG° values). The models amass 106 numerical RDKit descriptors, augmented with one-hot encodings of medium, site, radical, and structural class, and were evaluated through a leave-one-molecule-out protocol. Among the tested regression algorithms, the random forest regressor provides the best balance of accuracy and robustness with both R² test (≈0.94) and MAE (2.74 kcal mol⁻¹; RMSE (≈5.0 kcal mol⁻¹)), close to DFT chemical accuracy. The feature-importance analysis revealed that “electronic” and “experimental” (site/group) descriptors primarily drive predictions, with the radical’s maximum absolute partial charge being the most important descriptor in the prediction of a radical’s ΔG°. These results suggest that descriptor-driven RF (Random Forest) models can generalize across chemical space to provide interpretable ΔG° predictions, providing a path for chemists towards a scalable route to prioritize antioxidant candidates for broader molecular families. Full article

How non-dominant specialized firms sustain systemic innovation influence in modular service ecosystems without occupying architectural control positions remains theoretically underdeveloped. This study develops a dual-path Resource–Strategy–Capability (R–S–C) mechanism framework to explain how structurally distinct network positions generate divergent innovation trajectories among non-dominant firms. The empirical analysis draws on large-scale patent collaboration network data from China’s autonomous driving industry, covering 26 hidden champion firms and 14 global leading enterprises across 2009–2023. The framework identifies two divergent pathways: firms occupying structural hole positions adopt specialization-deepening strategies that build module-anchoring capabilities, while firms with high betweenness centrality adopt T-shaped strategies that build interface-bridging capabilities—both enabling systemic influence without architectural control. To make the resource construct theoretically precise, the framework distinguishes four categories of network-derived resources operative in the R–S–C mechanism—informational, coordination, reputational, and module-definition resources—and specifies three microfoundational processes through which strategic orientation translates into capability: experiential learning, codification of routines, and legitimation through external recognition. Institutional policy environments moderate these mechanisms by reshaping network structural heterogeneity rather than directly driving firm outcomes. The study challenges the canonical prediction of structural hole theory by demonstrating that brokerage positions generate specialization deepening rather than scope expansion when absorptive capacity constraints are binding, extends service ecosystem theory by introducing non-dominant firm pathways to systemic value co-creation, and reframes institutional policy as a network-structural moderator with transferable implications beyond the Chinese context. Full article

The rapid evolution of intelligent cluster systems—such as UAV swarms and networked autonomous agents—has brought spatial optimization and decision-making to the forefront of intelligent systems research. This paper provides a systematic and critical review of recent advances in spatial optimization for multi-agent intelligent clusters, focusing on four core domains: UAV swarm path planning, resource allocation, traffic network analysis, and visualization technologies. A bibliometric analysis based on the Web of Science Core Collection (2000–2024) identifies two major methodological transitions. In path planning, research has moved from traditional algorithms (A*, Dijkstra, dynamic programming), effective in static settings but limited in dynamic and large-scale applications, to bio-inspired optimization and deep reinforcement learning methods that improve adaptability and efficiency. In resource allocation, studies have shifted from centralized single-algorithm models to distributed, self-organizing hybrid frameworks that enhance robustness and real-time responsiveness. Moreover, intelligent cluster technologies are increasingly applied to urban traffic management and visualization, where analysis has advanced from static 2D mapping to interactive 3D and immersive VR/AR environments. A comparative framework is proposed to evaluate existing algorithms by adaptability, computational complexity, and scalability. The review concludes that future research should emphasize hybrid algorithm integration, cross-disciplinary data-driven modeling, and immersive visualization to support real-time decision-making. This study consolidates the evolutionary trajectory of intelligent cluster optimization, identifies critical research gaps, and outlines a roadmap for the next generation of intelligent spatial optimization systems. Full article

Learning path optimization is crucial in intelligent educational systems, with the core challenge of efficient multi-objective sequential decision-making under complex prerequisite constraints. To address the poor generalization of existing methods relying on fixed operator scheduling or handcrafted heuristics, this paper proposes a hyper-heuristic framework based on Dueling Deep Q-Network (Dueling DQN-HH), formulating operator selection as a sequential decision-making process for dynamic adaptive scheduling of low-level operators. The framework adopts priority-based encoding to unify learning path representation (decoupling the hyper-heuristic layer from the problem domain) and designs a composite reward mechanism integrating reward shaping, exploration incentives, and computational cost awareness to balance solution quality and efficiency. Additionally, it employs a dueling network architecture with prioritized experience replay to enhance policy learning stability. Experimental results show the proposed method outperforms representative baseline algorithms in solution quality, convergence stability, and computational efficiency. The framework demonstrates superior performance across multiple objectives, particularly in minimizing the total learning time (F_time), as validated on two heterogeneous datasets: MOOCCube (Computer Science) and PsyDataset (Psychology). Further ablation studies and operator evolution analyses verify its adaptive scheduling capability under different objectives and knowledge graph structures, demonstrating strong objective independence and cross-dataset generalization. Full article

Conservation management of large, multi-species landscapes requires integrating heterogeneous data streams—such as satellite imagery, GPS telemetry, camera traps, bioacoustic sensors, weather stations, and field reports—into a unified model capable of simulating ecosystem dynamics and generating actionable recommendations. This paper proposes a tiered, energy-aware AI architecture for constructing ecosystem digital twins that enables prescriptive, rather than merely descriptive or predictive, landscape-scale conservation management. The framework classifies conservation tasks across three computational tiers: classical machine learning for continuous environmental monitoring and species distribution prediction, deep learning for perception-oriented tasks such as computer vision and bioacoustic analysis, and foundation models for cross-domain synthesis and stakeholder interaction. We apply this architecture to a comprehensive digital twin of the Greater Yellowstone Ecosystem, anchored in the ongoing conservation crisis of the Sublette Pronghorn Herd—a population that crashed from 43,000 to 24,000 animals in a single winter due to compounding severe weather and a Mycoplasma bovis outbreak. We formalize a coupled change model linking population dynamics, forage condition, corridor permeability, winter severity, and disease pressure, and demonstrate how a prescriptive recommendations engine can generate goal-conditioned management actions for the herd’s 165-mile “Path of the Pronghorn” migration corridor. A comparative energy footprint analysis, grounded in hardware-level energy measurements using Intel RAPL instrumentation and the CodeCarbon framework, estimates that the tiered architecture reduces computational energy consumption by approximately 34% relative to a deep-learning-everywhere baseline and by over three orders of magnitude relative to a foundation-model-centric baseline. The architecture provides a replicable blueprint for resource-constrained conservation organizations seeking to deploy AI-powered ecosystem management at landscape scale. Full article

Stock price prediction is a challenging problem in quantitative investment, as financial markets generate complex, noisy, and dynamic time series containing heterogeneous signals. Short-term fluctuations usually exhibit greater uncertainty and stronger local variation, whereas long-term trends convey relatively stable and persistent information shaped by market and macroeconomic conditions. However, most existing methods struggle to distinguish these two components effectively, often leading to interference between short-term fluctuations and longer-term trends. In addition, they fail to capture dynamic temporal dependencies and cross-stock information propagation while preserving the causal structure of financial time series. To tackle these issues, we propose the Wavelet-Decoupled Spatiotemporal Network (WaveDSTN). It leverages wavelet transformation to decompose stock returns into high-frequency and low-frequency information, corresponding to short-term fluctuations and long-term trends, respectively. This decomposition enables the model to learn complementary predictive patterns more effectively. Furthermore, WaveDSTN incorporates a Dual-Path Spatiotemporal Encoder to capture complex temporal dependencies and evolving cross-stock information propagation while preserving temporal order and causal consistency. Extensive experiments demonstrate that WaveDSTN achieves significant improvements over existing methods, showing that explicitly modeling trend and fluctuation components can enhance predictive accuracy and reduce uncertainty in stock return forecasting. Full article

Product defect inspection is critical in industrial applications; however, it remains increasingly challenging in mass production environments, particularly for glossy or curved surface products. Conventional inspection of such surfaces typically relies on manual visual examination using gauges and operator judgment, which is time consuming and prone to inconsistency. This study proposes a robust defect detection framework for curved and reflective surfaces using a KEYENCE displacement laser sensor. The system integrates the Dijkstra algorithm, the Nearest Neighbor Algorithm, and a Genetic Algorithm to optimize the laser scanning path for structured image data generation. To validate the proposed framework, datasets were generated from both healthy and defective samples and used to train multiple deep learning models. A comparative analysis was conducted using YOLOv8, YOLOv9, YOLOv10, and YOLOv11 architectures. Experimental results demonstrate that YOLOv11 achieved the best overall performance, attaining an mAP50 score of 0.844 while also exhibiting lower computational complexity and faster inference. Full article

Post-disaster rescue scenarios often involve complex and variable terrains, imposing heterogeneous mobility requirements on different transport modes. Single-type vehicles face challenges in independently completing comprehensive rescue tasks. This study addresses the critical problem of coordinating heterogeneous aerial and ground vehicles to collaboratively plan relay rescue routes. To tackle the NP hard multi-terrain, multi-vehicle, and multi-route path planning problem, we propose a New Adaptive Multi-Strategy Particle Swarm Optimization algorithm (AMS-PSO-NEW). The algorithm features a synergistic integration of differential evolution’s multi-strategy mutation, SHADE-based adaptive parameter control, population diversity monitoring with restart mechanisms, and multi-level local search. A sequential hybrid mechanism is designed in which DE-generated trial vectors serve as reference positions for PSO velocity updates, enabling balanced global exploration and local exploitation. By leveraging adaptive parameter tuning, success history memory, and diverse population maintenance, AMS-PSO-NEW effectively overcomes premature convergence and low accuracy issues typical in discrete combinatorial optimization using traditional PSO, achieving a balanced global exploration and local exploitation. Performance validation is conducted over six rescue scenarios varying in scale and complexity, benchmarking AMS-PSO-NEW against nine algorithms: PSO, GA, NSGA-II, GWO, DE, ABC, CS, Q-learning, and MIP. Results demonstrate superior performance across four metrics (rescue success rate, average rescue time, total cost, and fairness), with significant improvements in high-complexity environments. Full article

High-precision path tracking is crucial for the safe operation of autonomous vehicles. However, the performance of Model Predictive Control (MPC) depends heavily on the accuracy of the vehicle dynamics model, whereas Deep Reinforcement Learning (DRL) generally lacks formal safety guarantees during training and exploration. To address this issue, this paper proposes a hybrid path-tracking framework, termed CRL-MPC, which integrates High-Order Control Barrier Function (HOCBF)-based reinforcement learning feedforward control with model predictive feedback control. Specifically, a Deep Deterministic Policy Gradient (DDPG) agent generates nominal feedforward steering commands, which are then corrected online by a High-Order Control Barrier Function (HOCBF)-based safety filter through a Quadratic Programming (QP) problem. During training on a high-fidelity CarSim–Simulink–Python co-simulation platform, the HOCBF-based safety filter constrains exploration within physically feasible regions, thereby preventing simulator failure caused by dynamically unsafe actions and improving training stability and sample efficiency. Meanwhile, the MPC controller provides feedback correction to compensate for residual errors. Comparative simulations were conducted against two baseline architectures: a standalone conventional MPC controller and a reinforcement-learning-based MPC(RL-MPC) hybrid architecture without the HOCBF-based safety filter. The results show that CRL-MPC achieves superior overall performance in path-tracking accuracy, control smoothness, and lateral dynamic stability. Compared with conventional MPC, CRL-MPC reduces the maximum lateral displacement error and its root mean square error (RMSE) by 54.1% and 62.7%, respectively, and reduces the maximum heading-angle error and its RMSE by 18.1% and 27.1%, respectively. Full article

The sharp dose gradients that underpin the dosimetric advantage of particle therapy over photon therapy can be undermined by the interplay effects due to intra-fraction motion in modern pencil beam scanning systems. Fluoroscopy-Guided Particle Therapy (FGPT) offers a promising path to improved motion management through real-time tracking of tumors or surrogate signals. The advent of flat-panel detector (FPD)-based technology has enabled tighter integration of fluoroscopy/fluorography into treatment units and accelerated clinical adoption and research, with commercial systems such as Hitachi’s Real-time Gated Particle Therapy (RGPT) now available. However, the need for implanted fiducial markers, with the associated invasiveness and risk of complications, limits the utility of RGPT to a few anatomic sites in selected patients. The full potential of FGPT, therefore, depends on reliable marker-less tumor tracking, which remains challenging because soft-tissue targets are obscured by overlapping anatomy along the X-ray path, leading to reduced reliability of traditional image-registration algorithms in the projection domain. Recent advances in deep learning and AI-driven image registration have renewed hope for overcoming these barriers, enabling real-time marker-less tracking for particle therapy. This review outlines the evolution of fluoroscopy technology from image intensifier (II) to FPD-based systems, summarizes historical and recent vendor-supported FGPT strategies, and surveys emerging AI-based algorithms in the literature. A general review of machine learning-based image registration is provided, challenges in generalizability and interpretability are highlighted, and potential paths toward reliable, clinically deployable FGPT are discussed. Full article

Unmanned Aerial Vehicles (UAVs) have become a key technology in precision agriculture, enabling efficient monitoring, inspection, and targeted interventions. However, effective UAV path planning in such environments requires the generation of safe, energy-efficient, and smooth trajectories in complex three-dimensional spaces. This paper proposes an Adaptive Q-Learning Guided Gorilla Troops Optimizer (AQGTO) for 3D UAV path planning. The proposed method integrates a state-aware Q-learning mechanism into the Gorilla Troops Optimizer (GTO), enabling the optimizer to adaptively select exploration, exploitation, and diversification strategies according to the current optimization state. A multi-objective cost function is formulated to simultaneously minimize path length, an energy-related surrogate cost, obstacle proximity, path smoothness, and altitude variation. In addition, a feasibility repair mechanism is introduced to ensure collision-free trajectories in environments with cylindrical obstacles. The proposed approach is evaluated in three representative agricultural scenarios: row-crop fields, orchard environments, and hilly terrains. Experimental results show that AQGTO achieves competitive and improved performance compared with classical A*, Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Whale Optimization Algorithm (WOA), and the original GTO in terms of trajectory cost, path efficiency, and stability. Furthermore, an ablation study confirms that the integration of Q-learning significantly enhances optimization performance. These results suggest that AQGTO provides an effective and robust solution for UAV path planning in complex agricultural environments. Full article

Integrated die-cast components reduce machining/assembly steps and improve mechanical dynamic characteristics, eliminating joint loosening/fracture risks after long-term use. However, the highly variable geometries and random spatial distributions of burrs, flash, parting lines, and risers in castings invalidate pre-programmed or teach-in robotic grinding methods. This paper reviews recent progress and future trends in robotic grinding, analyzing four core aspects: force control stability/adaptability (e.g., adaptive impedance control can reduce average force-tracking error to 0.38 N), trajectory planning/path generation (e.g., error-driven compensation can lower contour error by 34.2–55.1%), process parameter optimization, and challenges of sensing latency/quality evaluation (e.g., deep learning models achieve 97.64% accuracy in identifying abrasive belt wear states). The key enabling technologies are summarized, including active/passive compliant force control, model-/data-driven adaptive trajectory planning, intelligent process parameter optimization integrating physical mechanisms and data-driven approaches, and multi-modal state monitoring with online quality assessment. Representative applications (metal castings, aero-engine blades, thin-walled components, weld seams) are presented, and prospective research directions are proposed. This paper provides a comprehensive reference for theoretical research and engineering practice in this field. Full article

Existing research on cooperative control of connected and autonomous vehicles (CAVs) has primarily focused on structured freeway environments. Most existing approaches adopt lane-based modeling and discrete lane-change actions. These assumptions are unsuitable for toll plaza diverging areas without lane markings, where vehicles move toward multiple tollbooths. The absence of predefined lanes leads to continuous trajectory evolution, dense interactions, and increased safety risk. To address this limitation, this study proposes a multi-agent cooperative control framework based on Multi-Agent Proximal Policy Optimization (MAPPO) under a Centralized Training and Decentralized Execution (CTDE) architecture. The multi-agent formulation captures multi-vehicle interaction in toll plaza diverging areas, while centralized training improves learning stability. A target-path-oriented action space is introduced to replace the discrete lane-change action, enabling flexible tollbooth selection and continuous trajectory generation. The proposed cooperative strategy is trained and evaluated on a simulation platform structured under a Perception-Decision-Action framework, which provides a high-fidelity environment for weak-constraint traffic interactions. Simulation results based on real-world traffic data show that the proposed method improves traffic efficiency and enhances collision avoidance. Furthermore, comparative analyses are conducted to evaluate the model performance under varying traffic environments. Full article

Driven by the global energy transition and carbon neutrality goals, hybrid power systems have become a core technical path for energy conservation and carbon reduction in the transportation and power sectors, and the performance of energy management strategies directly determines the system’s overall energy efficiency. Traditional energy management methods have inherent bottlenecks of high model dependence and poor adaptability, making it difficult to satisfy real-time decision-making requirements under complex operating conditions. Deep Reinforcement Learning (DRL) provides an innovative solution to this technical bottleneck, and has become a cutting-edge research direction in this field. However, existing reviews have not yet constructed a full-chain analysis framework covering its algorithms, applications, verification, challenges and prospects. Focusing on the engineering application of DRL in the real-time energy management of hybrid power systems, this paper systematically sorts out domestic and international research results up to the first quarter of 2026. The core quantitative findings of this review are as follows: (1) DRL-based strategies can achieve 93–99.5% of the Dynamic Programming (DP) theoretical global optimum in fuel economy, which is 5–25% higher than rule-based methods; (2) DRL strategies only have 3.1–4.8% performance degradation under unseen operating conditions, which is significantly better than the 10.3–14.7% degradation of the Equivalent Consumption Minimization Strategy (ECMS); (3) Actor–Critic (AC) algorithms (Twin Delayed Deep Deterministic Policy Gradient (TD3)/Soft Actor–Critic (SAC)) have become the mainstream in this field, with a 3–5 times higher sample efficiency than value function-based algorithms; and (4) offline DRL and transfer learning can reduce the training time of DRL strategies by more than 80% while maintaining equivalent optimization performance. This paper first analyzes the essential attributes and core technical challenges of hybrid power system energy management; second, classifies DRL algorithms from the perspective of control engineering and analyzes their technical characteristics; third, disassembles the application design logic of DRL around four major scenarios: land vehicles, water vessels, aerial vehicles and fixed microgrids; fourth, summarizes the mainstream verification platforms and evaluation systems; fifth, analyzes core bottlenecks and cutting-edge solutions; and finally, prospects the development trends of next-generation intelligent energy management systems combined with cross-fusion technologies. This paper aims to build a complete technical system map for this field and promote the engineering deployment and practical application of intelligent energy management technologies integrating data and knowledge. Full article

The exhaust gas temperature (EGT) of the gas generator is a critical indicator for the health management system of a marine gas turbine engine. Therefore, EGT prediction can not only support predictive maintenance decision-making but also serves as a reliable virtual sensor for EGT measurement. However, the engine EGT exhibits strongly nonlinear coupling relationships with other gas path variables, which causes challenges for data-driven prediction. Graph neural networks (GNNs) are particularly effective in capturing the coupling relationships among gas path sensor variables. However, conventional static graph structures fail to characterize the varying coupling strengths under different operating conditions. In this study, a thermodynamic knowledge-driven graph attention network (TKD-GAT) method is proposed for accurate and robust EGT prediction. First, a physics-guided graph topology is constructed based on the gas turbine thermodynamic equations. Subsequently, a multi-head attention mechanism is introduced to generate edge weights that capture the varying thermodynamic coupling strengths under different operation conditions. The proposed model is evaluated on a real-world LM2500 gas turbine, which is widely used in modern propulsion systems of commercial and military ships. The ablation study confirms that the thermodynamic knowledge-driven graph topology and the attention mechanism-based edge weights are both necessary to enhance the EGT prediction performance. The TKD-GAT model shows the best performance with an RMSE of 0.446% and an R² of 0.971 compared with state-of-the-art models. The paired t-test and effect size measurement (Cohen’s d) statistically confirm the significance of performance improvements. The statistical results from multiple independent experiments prove the stability of the TKD-GAT model. Additionally, the model achieves a competitive computational cost despite the integration of a physics-guided graph topology and attention mechanisms. Crucially, an interpretability analysis confirms that the learned attention weights adhere to thermodynamic principles under different operation conditions. The proposed TKD-GAT model provides an effective solution for EGT prediction in health management systems. Full article