Search Results (865)

In response to the redundant spatio-temporal modeling and insufficient adaptation to dynamic decision-making in eye-tracking interaction of smart home interfaces, a smart home interface eye-tracking response optimization model based on spatio-temporal Transformer and gate control cross-attention is proposed. It adapts the physiological characteristics of eye-tracking jumps through dynamic sparse attention gating to compress computational redundancy and combines multi-objective reinforcement learning attention modulation to construct a closed-loop decision-making mechanism, optimizing interface parameters in real-time. Experiments showed that the model reduced eye-tracking trajectory prediction error by 23.7% compared to advanced benchmarks, increased the success rate of adapting to dynamic mutation scenarios to 89.2%, and controlled performance fluctuations within 2.3% under noise interference. In high-fidelity user testing, the accuracy of cross-task gaze transfer reached 93.4%, the failure rate of glare interference was optimized to 2.4%, and the user cognitive load index was reduced by 27.9%. Its resource consumption and energy consumption were reduced by 26.7% and 44.9%, respectively, while its posture deviation tolerance remained at 3.5°. The sparse spatio-temporal modeling of the spatio-temporal adaptive Transformer module and the enhanced gating mechanism of the hierarchical gated cross-attention module work together to break through the limitations of traditional methods in computational efficiency and dynamic feedback, providing high-precision and low-latency eye-tracking interaction solutions for smart home interface systems, and promoting the practical evolution of personalized human–machine collaborative control. Full article

Against the dual backdrop of worsening air pollution and industrial intelligent transformation, industrial robot technology has become an important means to promote air pollution synergistic governance. This study innovatively incorporates public environmental satisfaction and industrial robot application as dynamic mechanism variables, constructing an evolutionary game model involving the government, industrial enterprises, and the public. Through theoretical analysis and numerical simulation, the study reveals the influence mechanism of key cost–benefit parameters on stakeholders’ strategic interaction and the system’s evolution path. The conclusions are as follows: (1) The government’s environmental supervision directly affects enterprises’ green transformation willingness, and enterprises’ behavior reversely impacts public satisfaction and supervision effectiveness, forming a “supervision–response–feedback” closed-loop. (2) The cost and benefit parameters related to industrial robots are crucial for the evolution of the game system, and there is significant heterogeneity in their impact on the strategic choices of the three parties. The robot adaptation transformation of enterprise industrial depends on the comprehensive consideration of the transformation cost and the green benefits. Public supervision is regulated by both the supervision cost and the incentive benefit. The government regulation takes into account both the regulatory cost and the loss of social reputation. Various parameters dynamically regulate the system’s equilibrium by altering the party’s cost–benefit structure. (3) The application of industrial robots and the feedback of public environmental satisfaction form a coupling effect, jointly determining the long-term evolution direction of the game system. When the cost benefit and supervision incentives are well-matched, enterprises will actively promote the green transformation of industrial robots in order to achieve intelligent pollution control. The effectiveness of public supervision has also been fully realized. The dynamic adaptation of the two components can lead the system towards an efficient and stable equilibrium in air pollution governance. Full article

Agricultural intelligent agent technology features autonomy in multimodal perception, scalability for cross-scenario collaboration and adaptability via closed-loop optimization, serving as a core technological pillar for industrial intelligent upgrading and refined production management. This paper systematically elucidates its technical essence and methodological framework, focusing on five key aspects: multimodal heterogeneous data perception and fusion, scenario-oriented knowledge modeling and dynamic memory, intelligent decision-making and planning, embodied artificial intelligence, and closed-loop feedback optimization. On this basis, the paper outlines its core agricultural applications in four domains: crop cultivation, efficient utilization of agricultural resources, intelligent upgrading of agricultural technologies and equipment, and collaborative governance of the entire agricultural industry chain. From an interdisciplinary “AI + Agriculture” perspective, the paper further analyzes its future development directions, aiming to provide insights for improving agricultural intelligent agent technologies and promoting their industrial application to accelerate agricultural intelligent transformation. This study constructs a three-dimensional integrated methodological framework encompassing technological analysis, application mapping and trend forecasting, systematically summarizes its agricultural application scenarios and technological evolution characteristics, enriches the theoretical system and methodological construction of agricultural intelligent agent research, and provides a reusable analytical paradigm for agricultural intelligent agent research and practice. Full article

In flexible DC transmission and AC-DC interconnection systems, the Self-Adaption Station and Line Commutation Converter (SLCC) integrates static var compensation with conventional thyristor conversion functionality. This enables dynamic reactive power support at the valve side while improving commutation conditions, thereby enhancing the voltage support capability and operational robustness of DC systems. Under high renewable energy penetration, power fluctuations and sudden ramping challenges principal–slave controlled SLCC DC interconnection systems with a trade-off between principal-side DC voltage regulation and capacity margin constraints: Disturbance-induced active power demands may exceed available margins, causing DC voltage deviations and increasing protection trip risks. Leveraging the active/reactive decoupling characteristics of the SLCC topology, this paper proposes a principal–slave coordinated control strategy that accounts for principal station capacity margins. Methodologically, capacity margins are explicitly embedded into the principal station control mode. By reconstructing key variables in the DC voltage outer loop and introducing a closed-loop suppression mechanism with “over-capacity power” as feedback, the principal station maintains continuous voltage regulation while avoiding entry into over-capacity operation zones. On the slave side, a power support mechanism is designed to coordinate regulation among generation, storage, and load under power balance and equipment capacity constraints. This coordination process is formulated as a multi-objective optimization problem balancing disturbance economic losses with generation/storage utilization, solved using NSGA-II. Simulation results demonstrate that this strategy suppresses the risk of principle station overcapacity, enhances power sharing coordination during disturbance conditions, and improves DC voltage dynamic performance. Full article

This paper proposes a collaborative energy management strategy based on power prediction and deep reinforcement learning (DRL) to address the trade-offs among economic efficiency, durability, and dynamic performance in fuel cell hybrid power systems (FCHPS) under dynamic driving conditions. First, a hybrid prediction model termed LSTM-LSSVM with Cascade Correction (LSTM-LSSVM-CC) is developed. The cascade correction (CC) mechanism adopts a hierarchical structure to capture both low-frequency steady-state trends and high-frequency dynamic fluctuations, which are typically challenging for single models to represent. By integrating an online residual correction mechanism, this model generates accurate future power demand sequences. Second, a Dynamic Spatio-Temporal Fusion (DSTF) method is introduced to construct a high-dimensional DRL state space. This approach integrates predicted data, historical residuals, and real-time system states, enabling the agent to perform anticipatory decision-making. Third, a Dynamic Hierarchical Adaptive Multi-Objective Optimization Framework (DHAMOF) is designed. This framework dynamically adjusts objective weights and constraint boundaries based on real-time operating characteristics, enabling adaptive switching of optimization priorities across diverse scenarios. Furthermore, a closed-loop control architecture comprising “prediction–decision–execution–feedback” is established. By incorporating rolling horizon optimization and a proportional-integral (PI) residual compensation mechanism, the proposed architecture effectively suppresses prediction error accumulation and mitigates communication delays. Simulation results under combined CLTC-P and WLTP driving cycles demonstrate that, compared to conventional fixed-weight strategies, the proposed method achieves an 11.3% reduction in hydrogen consumption, a 30.9% decrease in $SOC$ fluctuation range, and a 55.3% reduction in power tracking error. Moreover, under disturbance scenarios involving prediction errors, sensor noise, and a 200 ms communication delay, the system exhibits superior robustness: the increase in hydrogen consumption is limited to within 8.3 g/100 km, and the power tracking error is reduced by 65.6% relative to uncorrected baselines. This collaborative optimization approach overcomes the limitations of traditional open-loop prediction and fixed-weight control, offering a novel technical pathway for the high-efficiency and stable operation of fuel cell hybrid power systems. Full article

With the growing demand for trustworthy multi-party data sharing, federated learning has demonstrated broad potential in cross-entity collaborative modeling. However, it still faces challenges such as insufficient participant engagement, inaccurate contribution assessment, and the lack of dynamic profit-sharing mechanisms. Traditional incentive schemes, which typically rely on game-theoretic models or static rules, struggle to accommodate dynamic client participation and heterogeneous data distributions, thereby degrading the convergence efficiency and generalization performance of the global model. To address these issues, we propose a budget-aware closed-loop incentive allocation for federated learning with deep deterministic policy gradient (DDPG). The proposed approach constructs a DDPG-driven closed-loop framework in which the server manages system states, incentive decisions, and model aggregation, while clients autonomously adjust their data contribution levels. By formulating incentive allocation as a sequential decision-making problem, the mechanism jointly optimizes policy and value functions. A permutation method is introduced to ensure invariance to client ordering, and an Ornstein–Uhlenbeck process is employed to enhance exploration, thereby improving the adaptiveness and overall effectiveness of incentive allocation. Experimental results show that the proposed method significantly increases cumulative rewards and improves client data-sharing rates in high-dimensional dynamic environments. Compared with traditional fixed incentive schemes, the mechanism demonstrates clear advantages in adaptiveness, incentive effectiveness, and model performance. Full article

Bio-inspired quadrupedal robots exhibit superior adaptability and mobility in unstructured environments, making them suitable for complex task scenarios such as navigation, obstacle avoidance, and tracking in a variety of environments. Visual perception plays a critical role in enabling autonomous behavior, offering a cost-effective alternative to multi-sensor systems. This paper proposes an image segmentation-guided visual tracking framework to enhance both perception and motion control in quadruped robots. On the perception side, a cascaded convolutional neural network is introduced, integrating a global information guidance module to fuse low-level textures and high-level semantic features. This architecture effectively addresses limitations in single-scale feature extraction and improves segmentation accuracy under visually degraded conditions. On the control side, segmentation outputs are embedded into a biologically inspired central pattern generator (CPG), enabling coordinated generation of limb and spinal trajectories. This integration facilitates a closed-loop visual-motor system that adapts dynamically to environmental changes. Experimental evaluations on benchmark image segmentation datasets and robotic locomotion tasks demonstrate that the proposed framework achieves enhanced segmentation precision and motion flexibility, outperforming existing methods. The results highlight the effectiveness of vision-guided control strategies and their potential for deployment in real-time robotic navigation. Full article

In nanoscale wafer defect inspection, raster scan imaging imposes sub-micrometer requirements on motion stage tracking accuracy, while trajectory changes and load variations pose significant challenges to traditional control methods. This paper proposes a Real-time Iterative Compensation based Adaptive Robust Control (RICARC) strategy. Within this framework, the ARC module incorporates RLS-based online parameter estimation, a PID-type feedback control term, and a robust control term to suppress lumped disturbances. On this basis, the RIC module establishes a discrete prediction model based on the ARC closed-loop system and iteratively generates optimal feedforward compensation signals at each sampling instant to further suppress residual tracking errors. Experimental results across five operating scenarios, including periodic, dual-frequency, and S-curve trajectories, as well as payload variation, and strong external disturbances, demonstrate that RICARC consistently achieves sub-micrometer RMS accuracy ranging from 0.120 to 0.240 μm, reducing RMS errors by over 75% compared with conventional ARC, effectively enhancing imaging quality in nanoscale wafer defect detection systems. Full article

Friction stir welding (FSW) is a critical solid-state joining process for lightweight and high-performance metallic structures, particularly in aerospace and automotive manufacturing, yet conventional implementations remain largely dependent on offline parameter optimization and open-loop control. The purpose of this review is to examine how Industry 4.0 technologies enable the transition of FSW from a parameter-driven process into an intelligent, adaptive, and increasingly autonomous manufacturing capability. A structured review methodology was employed, including systematic literature selection and synthesis of recent research on smart sensing, industrial internet of things (IIoT), data analytics, machine learning, digital twins, automation, robotics, and human–machine interaction in FSW. The review reveals that Industry 4.0 integration enables real-time process monitoring, predictive quality assurance, closed-loop control, and virtual process optimization, resulting in improved weld quality, reliability, productivity, and scalability. Significant benefits are observed for safety-critical aerospace components and high-throughput automotive production, where adaptability and consistency are essential. However, persistent challenges remain in data standardization, model generalization, real-time digital twin integration, interoperability, cybersecurity, and workforce readiness. This review concludes that addressing these challenges through interdisciplinary research, standardization efforts, and human-centered system design is essential for enabling adaptive and data-driven FSW systems. The findings position intelligent FSW as a foundational technology for smart, resilient, and sustainable metal manufacturing in the Industry 4.0 era. Full article

Continuous impact vibrations caused by uneven road surfaces (such as speed bumps) can significantly reduce the trajectory tracking accuracy of mobile manipulator. This study proposes for the first time an integrated framework combining a semi-active magnetorheological elastomer (MRE) intelligent isolation system with an active trajectory tracking controller to improve the operational accuracy of mobile manipulator under continuous impact excitation, and numerically evaluates the effect of the MRE isolation system. The working principle and design method of the MRE isolation system for mobile manipulators are described, and a multi-layer MRE isolator is fabricated and experimentally characterized. A semi-active control strategy is developed to adaptively adjust the stiffness and damping of the isolator based on continuous impact input. To further compensate for residual disturbances transmitted through the isolator, an enhanced computational torque control (CTC) and proportional-derivative (PD) controller with predefined-time disturbance observer (DOB) is designed for the mobile manipulator. This ensures that the disturbance estimate converges within a predefined time window, thereby improving the robustness of the closed-loop system. By constructing a comprehensive multibody dynamics model coupling the vehicle, the MRE isolator, and the manipulator, vibration transmission is analyzed and trajectory tracking performance is evaluated. Simulation results under continuous road impact excitation demonstrate that the proposed semi-active MRE intelligent isolation system can significantly suppress base vibration and greatly improve the trajectory tracking accuracy of the mobile manipulator end-effector and its joints. This study proves the feasibility of the semi-active MRE isolation system in the trajectory tracking application of mobile manipulator and provides a new approach for the collaborative design of intelligent vibration isolation and control strategies for mobile robot systems operating in harsh and frequently impacted environments. Full article

Engineering Materials courses are characterized by dense conceptual content, cumulative knowledge structures, and heterogeneous student learning trajectories. Existing teaching reform studies often focus on isolated instructional techniques or digital tools, while paying limited attention to the systemic organization of learning activities, assessment, feedback, and instructional decision-making. This study proposes a system-oriented teaching framework for an undergraduate Engineering Materials course within an urban underground space engineering program. The framework conceptualizes course instruction as a closed-loop process driven by continuous learning evidence and feedback regulation. The framework was implemented in an undergraduate Engineering Materials course with 50 students over a 16-week semester using a learning management platform. Multiple sources of process data were collected, including platform access records, assignment submissions, weekly quiz performance, pre- and post-course concept assessments, instructor feedback logs, and instructional adjustment records. The results indicate that the proposed framework supported timely instructional regulation and adaptive responses to heterogeneous learning states. Observable improvements were found in student engagement patterns and assessment outcomes across the semester. Mean concept test scores increased from 55.7 to 72.2. Students with lower initial scores gained an average of 22.3 points, compared to 11.8 points for their higher-performing peers. A total of 312 feedback messages were delivered, with a median latency of three days. These improvements were observed in association with the implementation of the framework, although causal attribution is limited by the non-experimental, single-cohort design. The study provides an exploratory case showing that system-oriented teaching design may offer a coherent and practically feasible approach for enhancing engineering education in data-rich instructional environments, while also contributing to the application of systems thinking in teaching reform. Full article

In light of the global issue of “Carbon Neutrality”, a high proportion of renewable energy integrated into modern power systems has become the key to energy strategic transformation, which has escalated the demand for high-gain, high-power converters for DC energy conversion. In this paper, a non-isolated double-layer multi-port parallel-connected high-gain DC–DC conversion system has been proposed. The system consists of two energy layers: the upper layer is designed as a non-isolated high-gain three-port DC conversion topology, which includes two energy inputs and one output port, and the bottom layer is a three-port constant current output module. The output ports of these layers are connected in parallel, while the input ports are independent. Thus, both high output voltage gain and power capacity were fulfilled for the renewable power application condition. The system is capable of operating in both input-parallel–output-parallel (IPOP) and multi-input–independent-output-parallel (MIIOP) modes, thereby enabling multi-port high-gain DC power conversion. Detailed analysis of the operation strategies under a switching cycle for both energy layers is presented. A small signal was introduced to establish the mathematical model of both energy topologies. In order to simultaneously regulate the output voltage and achieve dynamic current sharing between the layers, an adaptive current-sharing control strategy was developed based on the established system models. The proposed control strategy can control the output voltage through the upper-layer topology and dynamically allocates output current between the layers based on the output power level, which will effectively enhance the system’s power rating. The simulation mode was built in the PSIM environment, open-loop simulations were carried out for obtaining system characteristics, and closed-loop simulations were conducted for control efficiency validation. Finally, a 2000-W experimental prototype was developed based on the digital control center dsPIC33FJ64GS606. Open-loop and closed-loop experiments were carried out for system performance evaluation. Both simulation and experimental results successfully evaluated the power transfer performance and control system performance of the proposed system, and a peak efficiency of 95.7% under 10 times voltage gain was achieved. Full article

The transition toward sustainable energy systems is increasing the operational complexity of modern power grids due to the high penetration of renewable energy sources, distributed energy resources, and bidirectional energy flows. Artificial intelligence has emerged as a key enabling technology for forecasting, optimization, and control in smart grids. Current AI implementations in energy systems lack unified workflows integrating forecasting, decision-making, adaptive stability regulation, and distributed coordination. Moreover, existing control approaches rarely incorporate biologically inspired stability mechanisms such as homeostatic regulation, limiting system-level resilience under dynamic operating conditions. This work aims to develop an architectural framework in the form of a unified artificial intelligence meta-pipeline enabling homeostatic control and distributed resource optimization in sustainable energy systems through closed-loop intelligent operation. A layered artificial intelligence meta-pipeline architecture is proposed integrating system representation, data intelligence, decision intelligence, homeostatic feedback regulation, and distributed coordination. A formal Homeostatic Energy Index is introduced to quantify system stress and enable supervisory adaptive policy regulation. The framework is validated using a reproducible microgrid-level simulation combining reinforcement learning-based control with homeostatic feedback regulation. Experimental validation demonstrates stable closed-loop operation under stochastic demand and renewable variability. The framework maintains bounded system stress levels, achieving an average Homeostatic Energy Index of 18.17 while preserving near-zero energy imbalance performance, confirming that homeostatic feedback improves stability without degrading energy balancing performance. This work introduces a unified artificial intelligence meta-pipeline architectural framework and formally defines a homeostatic feedback layer for sustainable energy system control. The proposed approach enables stability-aware structured integration of heterogeneous AI components and provides a foundation for self-adaptive, resilient, and distributed intelligent energy systems. Full article

Global supply chains face intensifying pressures from disruption, regulatory complexity, and sustainability mandates, requiring a shift toward more resilient and adaptive coordination. While artificial intelligence (AI) and blockchain have been recognised as complementary enablers, their implementation remains largely fragmented, existing as isolated tools linked by manual data exchange rather than integrated, programmable logic. This paper addresses this orchestration gap by proposing the Dynamic Resource Orchestration Framework for AI-Blockchain Integrated Supply Chains (DROF-AIBC). Grounded in Resource Orchestration Theory (ROT) and Dynamic Capabilities Theory (DCT), the framework provides a theoretical foundation for the strategic and autonomous orchestration of digital resources. Unlike classic supply chain orchestration, which focuses on the linear coordination of physical assets and legacy systems, DROF-AIBC conceptualises an “intelligent conductor” as a coordination mechanism combining AI-driven analytics and smart contract-based execution. This mechanism supports the configuration, optimisation, and monitoring of resources in response to changing external signals, effectively closing the loop between analytical insights and verifiable execution. The paper further substantiates how this autonomous capability serves as a foundational roadmap for the Industry 5.0 paradigm, embedding human-centricity through Explainable AI (XAI) to provide a “provenance of logic”, promoting circular economy sustainability, and fostering systemic resilience in turbulent environments. The framework aims to provide both a theoretical foundation and a practical roadmap for orchestrating AI and blockchain to advance resilient, sustainable and adaptive supply chains. Full article

This article presents the synthesis, real-time implementation, and experimental validation of an approximated adaptive dynamic programming (AADP) actor–critic controller for precise flow rate regulation of a variable-displacement axial-piston pump designed for open-circuit hydraulic systems. Replacing the conventional hydro-mechanical regulator with an electrohydraulic proportional spool valve, the model-free controller employs two compact two-layer neural networks: the actor generates valve PWM signals from the flow tracking error, its integral, and measured discharge pressure, while the critic approximates the infinite-horizon quadratic cost-to-go via the online solution of the Bellman equation through gradient descent on Bellman residuals. Lyapunov analysis establishes closed-loop stability under bounded learning rates, with initial weights tuned via nominal plant simulation to ensure convergence from feasible starting policies. After extensive laboratory testing across four fixed loading conditions and dynamic load variations, the adaptive controller demonstrated superior performance compared with a proportional-integral (PI) controller, a Lyapunov model-reference adaptive controller (LMRAC), and an H_∞ controller (Hinf). Real-time metrics confirm bounded critic signals and near-zero Bellman errors, validating optimal policy convergence amid unmodeled hydraulic nonlinearities. Full article