Search Results (40,509)

Convolutional Neural Networks (CNNs) generally suffer from excessive computational overhead, high resource consumption, and complex network structures, which severely restrict the deployment on microprocessor chips. Existing related accelerators only have an energy efficiency ratio of 2.32–6.5925 GOPs/W, making it difficult to meet the low-power requirements of embedded application scenarios. To address these issues, this paper proposes a low-power and high-energy-efficiency CNN accelerator architecture based on a central processing unit (CPU) and an Application-Specific Integrated Circuit (ASIC) heterogeneous computing architecture, adopting an operator-fused systolic array algorithm with the YOLOv5n target detection network as the application benchmark. It integrates a 2D systolic array with Conv-BN fusion technology to achieve deep operator fusion of convolution, batch normalization and activation functions; optimizes the RISC-V core to reduce resource usage; and adopts a locking mechanism and a prefetching strategy for the asynchronous platform to ensure operational stability. Experiments on the Nexys Video development board show that the architecture achieves 20.6 GFLOPs of computational performance, 1.96 W of power consumption, and 10.46 GOPs/W of energy efficiency ratio, which is 58–350% higher than existing mainstream accelerators, thus demonstrating excellent potential for embedded deployment. Full article

Organic pollution poses a serious threat to global water safety, while traditional treatment technologies suffer from low efficiency, high costs, and secondary pollution issues. This study successfully develops a highly efficient separation and photocatalytic degradation composite bismuth oxychloride@graphene oxide/polyimide (BiOCl@GO/PI) membrane by loading GO and BiOCl photocatalysts onto PI supporting membrane. The results show that this composite membrane achieves a rejection of 99.8% for methylene blue (MB) and 87.6% for tetracycline hydrochloride (TC). Under UV irradiation, the membrane exhibits a retention rate decline of only 6.8% after five cycles, with water flux stably maintaining at 605 L m⁻² h⁻¹ bar⁻¹. Compared to dark conditions, it demonstrates remarkable flux recovery. This is attributed to the membrane’s excellent photocatalytic degradation activity under UV irradiation. After five degradation cycles, the degradation efficiency is decreased from 97.5 to 88.3%. Studies on radical scavengers indicate that UV irradiation generates free radicals, thereby conferring excellent catalytic activity to the membrane. Its unique synergistic effect between separation and photocatalysis endows it with outstanding self-cleaning performance. This research provides an innovative integrated solution for antibiotic pollution control, demonstrating significant potential for environmental applications. Full article

High electrostrain, excellent thermal stability, and low hysteresis are critical requirements for advanced high-precision actuators. However, simultaneously achieving these synergistic properties in lead-free ferroelectric ceramics remains a significant challenge. In this work, a targeted B-site doping strategy was employed to develop novel lead-free (0.99-x)BaTiO₃-xBaZrO₃-0.01Bi(Zn_2/3Nb_1/3)O₃ (BT-xBZ-BZN, x = 0–0.2) ceramics. Systematic investigation identified optimal Zr⁴⁺ substitution at x = 0.1, which yielded an outstanding combination of electromechanical properties. For this optimal composition, a high unipolar electrostrain (S_max = 0.11%) was achieved at 50 kV/cm, accompanied by an ultra-low hysteresis (H_S = 1.9%). Concurrently, a large electrostrictive coefficient (Q₃₃ = 0.0405 m⁴/C²) was determined, demonstrating excellent thermal robustness with less than 10% variation across a broad temperature range of 30–120 °C. This superior comprehensive performance is attributed to a composition-driven evolution from a long-range ferroelectric to a pseudocubic relaxor state. In this state, the dominant electrostrictive effect, propelled by reversible dynamics of polar nanoregions (PNRs), minimizes irreversible domain switching. These findings not only present BT-xBZ-BZN (x = 0.1) as a highly promising lead-free candidate for high-precision, low-loss actuator devices, but also provide a viable design strategy for developing high-performance electrostrictive materials with synergistic large strain and superior thermal stability. Full article

During batch soldering, assembly of micro image sensor modules, initial random pose, and feature partially occlude target micro-component image, leading to issues of missed and erroneous detection, and low 3D spatial positioning accuracy due to cross-depth-of-field detection errors in microscopic vision. This paper proposes Small object-YOLO11-Cross-Depth-of-field Positioning (SO-YOLO11-CDP), an instance segmentation-based approach for precision cross-depth-of-field positioning micro-component. First, an improved Small object-YOLO11 (SO-YOLO11) image segmentation algorithm is designed. By incorporating a coordinate attention mechanism (CA) into segmentation head to enhance localization of micro-targets, the backbone uses non-stride convolution to preserve fine-grained feature, while target regression performance is boosted via Efficient-IoU (EIoU) loss combined with normalized Wasserstein distance (NWD). Subsequently, to further improve spatial position detection accuracy in cross-depth-of-field detection, a calibration error compensation model for image Jacobian matrix is established based on pinhole imaging principles. Experimental results indicate that SO-YOLO11 achieves 16.1% increase in precision, 4.0% increase in recall, and 9.9% increase in mean average precision (mAP0.5) over baseline YOLO11. Furthermore, it accomplishes spatial detection accuracy superior to 6.5 μm for target micro-components. The method presented in this paper holds significant engineering application value for high-precision spatial position detection of micro image sensor components. Full article

Addressing the challenges of low resource-use efficiency and supply–demand mismatch in Hami melon production, this study investigated the interactive effects of irrigation and fertilization to identify an optimal regime that balances yield, water conservation, and resource-use efficiency (i.e., water use efficiency and fertilizer partial factor productivity). A greenhouse experiment was conducted in Hami, Xinjiang, employing a two-factor design with five irrigation levels (W1–W5: 60–100% of full irrigation) and three fertilization levels (F1–F3: 80–100% of standard rate), replicated three times. Growth parameters, yield, water use efficiency (WUE), and partial factor productivity of fertilizer (PFP) were evaluated and comprehensively analyzed using the entropy-weighted TOPSIS method, regression analysis, and the NSGA-II multi-objective genetic algorithm. Results demonstrated that irrigation volume was the dominant factor influencing growth and yield. The W4F3 treatment (90% irrigation with 100% fertilization) achieved the optimal outcome, yielding 75.74 t ha⁻¹—a 9.71% increase over the control—while simultaneously enhancing WUE and PFP. Both the entropy-weighted TOPSIS evaluation (C = 0.998) and regression analysis (optimal irrigation level at w = 0.79, ~90% of full irrigation) identified W4F3 as superior. NSGA-II optimization further validated this, generating Pareto-optimal solutions highly consistent with the experimental optimum. The model-predicted optimal regime for greenhouse Hami melon in Xinjiang is an irrigation amount of 3276 m³ ha⁻¹ and a fertilizer application rate of 814.8 kg ha⁻¹. This regime facilitates a 10% reduction in irrigation water and a 5% reduction in fertilizer input without compromising yield, alongside significantly improved resource-use efficiencies. Full article

Early fault vibration signals of rolling bearings are non-stationary and nonlinear, with weak fault signatures easily masked by noise. Traditional denoising methods (e.g., wavelet thresholding, empirical mode decomposition (EMD)) struggle to accurately extract effective features. Although variational mode decomposition (VMD) overcomes mode mixing, its core parameters rely on empirical selection, making it prone to local optima and limiting its denoising performance. To address this critical issue, this study aims to propose a hybrid model with adaptive parameter optimization and efficient denoising capabilities, enhancing the signal-to-noise ratio (SNR) and feature discriminability of early fault signals in rolling bearings. The novelty of this work is reflected in three aspects: (1) An improved edge strategy whale optimization algorithm (IEWOA) is proposed, incorporating six enhancements to balance global exploration and local exploitation. Using the minimum average envelope entropy as the objective function, the IEWOA achieves adaptive global optimization of VMD parameters. (2) A hybrid framework of “IEWOA-VMD + dataset-specific wavelet thresholding for secondary denoising” is constructed. The optimized VMD first decomposes signals to separate noise and effective components, followed by secondary denoising, ensuring both adaptable signal decomposition and precise denoising. (3) Comprehensive validation is conducted across five models using two public datasets (Case Western Reserve University (CWRU) and Paderborn Universität (PU)). Key findings demonstrate that the proposed method achieves a root-mean-square error (RMSE) as low as 0.00013–0.00041 and a Normalized Cross-Correlation (NCC) of 0.9689–0.9798, significantly outperforming EEMD, traditional VMD, and VMD optimized by single algorithms. The model effectively suppresses noise interference, preserves the fundamental and harmonic components of fault features, and exhibits strong robustness under different loads and fault types. This work provides an efficient and reliable signal preprocessing solution for early fault diagnosis of rolling bearings. Full article

Neurodegenerative diseases (NDDs), including Alzheimer’s disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD), are becoming more prevalent and still lack effective disease-modifying therapies (DMTs). However, translational efficiency remains critically low. For example, a ClinicalTrials.gov analysis of AD programs (2002–2012) estimated ~99.6% attrition, while PD programs (1999–2019) achieved an overall success rate of ~14.9%. In vitro platforms are assessed, ranging from immortalized neuronal lines and primary cultures to human-induced pluripotent stem cell (iPSC)-derived neurons/glia, neuron–glia co-cultures (including neuroinflammation paradigms), 3D spheroids, organoids, and blood–brain barrier (BBB)-on-chip systems. Complementary in vivo toxin, pharmacological, and genetic models are discussed for systems-level validation and central nervous system (CNS) exposure realism. The therapeutic synthesis focuses on AD, covering symptomatic drugs, anti-amyloid immunotherapies, tau-directed approaches, and repurposed drug classes that target metabolism, neuroinflammation, and network dysfunction. This review links experimental models to translational decision-making, focusing primarily on AD and providing a brief comparative context from other NDDs. It also covers emerging targeted protein degradation (PROTACs). Key priorities include neuroimmune/neurovascular human models, biomarker-anchored adaptive trials, mechanism-guided combination DMTs, and CNS PK/PD-driven development for brain-directed degraders. Full article

Micro-expression recognition (MER) is challenged by a brief duration, low intensity, and heterogeneous spatial frequency patterns. This study introduces a novel MER architecture that reduces computational cost by fine-tuning a large feature extraction model with LoRA, while integrating frequency-domain transformation and graph-based temporal modeling to minimize preprocessing requirements. A Spatial Frequency Adaptive (SFA) module decomposes high- and low-frequency information with dynamic weighting to enhance sensitivity to subtle facial texture variations. A Dynamic Graph Attention Temporal (DGAT) network models video frames as a graph, combining Graph Attention Networks and LSTM with frequency-guided attention for temporal feature fusion. Experiments on the SAMM, CASME II, and SMIC datasets demonstrate superior performance over existing methods. On the SAMM 5-class setting, the proposed approach achieves an unweighted F1 score (UF1) of 81.16% and an unweighted average recall (UAR) of 85.37%, outperforming the next best method by 0.96% and 2.27%, respectively. Full article

With the increasing integration of renewable energy sources into the grid, power quality at the point of common coupling (PCC)—particularly harmonic distortion introduced by power electronic converters—has become a critical concern. This paper presents a rigorous design and evaluation of a three-phase, fifteen-level cascaded H-bridge multilevel inverter (CHB MLI) with an LCL filter, selected for its superior harmonic attenuation, compact size, and cost-effectiveness compared to conventional passive filters. The proposed system employs Phase-Shifted Pulse Width Modulation (PS PWM) for balanced operation and low output distortion. A systematic, reproducible methodology is used to design the LCL filter, which is then tested across a wide range of switching frequencies (1–5 kHz) and grid impedance ratios (X/R = 2–9) in MATLAB/Simulink R2025a. Comprehensive simulations confirm that the filter effectively reduces both voltage and current total harmonic distortion (THD) to levels well below the 5% limit specified by IEEE 519, with optimal performance (0.53% current THD, 0.69% voltage THD) achieved at 3 kHz and X/R ≈ 5.6. The filter demonstrates robust performance regardless of grid conditions, making it a practical and scalable solution for modern renewable energy integration. These results, further supported by parametric validation and clear design guidelines, provide actionable insights for academic research and industrial deployment. Full article

The concept of sustainable development has emerged as a global consensus, forged in response to environmental constraints and critical reflection on conventional growth-oriented paradigms. It now serves as the overarching framework for addressing climate, ecological, and socio-economic crises. In the period after the adoption of the Sustainable Development Goals (SDGs) in 2016, there was an observable trend of increased integration of these objectives into the strategic frameworks of national and subnational entities. However, global assessments have indicated a divergence between the progress achieved and the trajectory delineated by the SDGs. The Earth system is demonstrating signs of decreased resilience, with widening inequalities and the emergence of multiple crises, thereby hindering the implementation of the 2030 Agenda for Sustainable Development. As the 2030 deadline approaches, a fundamental question arises for global development governance: what should be the future of the SDGs beyond 2030? While insufficient progress has prompted debates over the adequacy of the SDG framework, fundamentally revising or replacing the SDGs would risk undermining a hard-won international consensus forged through decades of negotiation and institutional investment. Based on a comprehensive review of the historical evolution of the sustainable development concept, this study argues that the SDGs represent a rare and fragile achievement in global governance. While insufficient progress has sparked debates about their effectiveness, fundamentally revising or replacing the SDGs would jeopardize the hard-won international consensus forged through decades of negotiations and institutional investments. This study further analyzes the latest progress on the SDGs and identifies emerging risks, aiming to explore how to accelerate and optimize sustainable development pathways within the existing SDG framework rather than propose a new global goal system. Based on both global experience and practice in China, four interconnected strategic priorities—namely, economic reform, social equity, environmental justice, and technology sharing—are proposed as a comprehensive framework to accelerate SDG implementation and guide the transformation of development pathways towards a more just, low-carbon, and resilient future. Full article

Ground-based full-physical experiments for space rendezvous and docking serve as a critical step in verifying the reliability of docking technology. The high-precision active attitude setting of spacecraft simulators represents a key technology for ground-based full-physical experiments. In order to satisfy the requirement for high-precision attitude control in these experiments, this paper proposes an enhanced method based on auto disturbance rejection control (ADRC). This paper addresses the limitations of traditional deadband–hysteresis relay controllers, which exhibit low steady-state accuracy and insufficient disturbance rejection capability. This approach employs a nonlinear extended state observer (NESO) to estimate and compensate for total system disturbances in real time. Concurrently, it incorporates an adaptive mechanism for deadband and hysteresis parameters, dynamically adjusting controller parameters based on disturbance estimates and attitude errors. This overcomes the trade-off between accuracy and power consumption that is inherent in fixed-parameter controllers. Furthermore, the method incorporates a nonlinear tracking differentiator (NTD) to schedule transitions, enabling rapid attitude settling without overshoot. The stability analysis demonstrates that the proposed controller achieves local asymptotic stability and global uniformly bounded convergence. The simulation results demonstrate that under three typical operating conditions (conventional attitude setting, pre-separation connector stabilisation, and docking initial condition establishment), the steady-state attitude error remains within ±0.01°, with convergence times under 3 s and no overshoot. These results closely match ground test data. This approach has been demonstrated to enhance the engineering applicability of the control system while ensuring high precision and robust performance. Full article

Amidst the ongoing global energy crisis, environmental deterioration, and the exacerbation of climate change, the development of renewable energy, particularly solar energy, has become a central topic in the global energy transition. This study investigates a solar photovoltaic thermal (PVT) heat pump system that utilizes an expanded honeycomb-channel PVT module to enhance the comprehensive utilization efficiency of solar energy. A simulation platform for the solar PVT heat pump system was established using Aspen Plus software (V12), and the system’s performance impact mechanisms and engineering applications were researched. The results indicate that solar irradiance and the circulating water temperature within the PVT module are the primary factors affecting system performance: for every 100 W/m² increase in solar irradiance, the coefficient of performance for heating (COP_h) increases by 13.7%, the thermoelectric comprehensive performance coefficient (COP_co) increases by 14.9%, and the electrical efficiency of the PVT array decreases by 0.05%; for every 1 °C increase in circulating water temperature, the COP_h and COP_co increase by 11.8% and 12.3%, respectively, and the electrical efficiency of the PVT array decreases by 0.03%. In practical application, the system achieves an annual heating capacity of 24,000 GJ and electricity generation of 1.1 million kWh, with average annual COP_h and COP_co values of 5.30 and 7.60, respectively. The Life Cycle Cost (LCC) is 13.2% lower than that of the air-source heat pump system, the dynamic investment payback period is 4–6 years, and the annual carbon emissions are reduced by 94.6%, demonstrating significant economic and environmental benefits. This research provides an effective solution for the efficient and comprehensive utilization of solar energy, utilizing the low-global-warming-potential refrigerant R290, and is particularly suitable for combined heat and power applications in regions with high solar irradiance. Full article

Human exposure to persistent organic pollutants such as polychlorinated biphenyls (PCB) and brominated flame retardants (BFR), including both legacy and emerging compounds, remains a concern due to their bioaccumulative nature and potential health effects. Comprehensive analytical methods are necessary to monitor these substances in complex biological matrices, such as human serum. A gas chromatography–mass spectrometry (GC-MS) method was developed for the simultaneous determination of 44 analytes, encompassing PCB and a broad spectrum of BFR with diverse physicochemical properties. The extraction procedure and GC-MS parameters were optimized using a design of experiments approach to maximize performance while minimizing analysis time. The method demonstrated high sensitivity, precision, and accuracy, thereby meeting internationally recognized validation criteria for biomonitoring applications. To further ensure analytical reliability, compound confirmation was achieved using gas chromatography–high-resolution mass spectrometry, providing enhanced selectivity and confidence in identification, particularly for low-level analytes. Key advantages of the method include its applicability to analytes with significantly different chemical behaviors and its capacity to quantify a large number of target compounds simultaneously. This makes it a powerful tool for assessing human exposure to both regulated and emerging halogenated contaminants. Full article

Covert communication assisted by unmanned aerial vehicles (UAVs) can achieve a low detection probability in complex environments through auxiliary strategies, including dynamic trajectory planning and power management, etc. This paper proposes a dual-UAV scheme, where one UAV transmits covert information while the other one generates stochastic jamming to disrupt the eavesdropper and reduce the probability of detection. We propose a dual-mode jamming scheme which can efficiently enhance the average covert rate (ACR). A joint optimization of the dual UAVs’ flight speeds, accelerations, transmit power, and trajectories is conducted to achieve the maximum ACR. Given the high complexity and non-convexity, we develop a dedicated algorithm to solve it. To be specific, the optimization is decomposed into three sub-problems, and we transform them into tractable convex forms using successive convex approximation (SCA). Numerical results verify the efficacy of dual-mode jamming in boosting ACR and confirm the effectiveness of this algorithm in enhancing CC performance. Full article

With the explosive growth in the quantity of electrical equipment in distribution networks, traditional manual inspection struggles to achieve comprehensive coverage due to limited manpower and low efficiency. This has led to frequent equipment failures including partial discharge, insulation aging, and poor contact. These issues seriously compromise the safe and stable operation of distribution networks. Real-time monitoring and defect identification of their operation status are critical to ensuring the safety and stability of power systems. Currently, commonly used methods for defect identification in distribution network electrical equipment mainly rely on single-image or voiceprint data features. These methods lack consideration of the complementarity and interleaved nature between image and voiceprint features, resulting in reduced identification accuracy and reliability. To address the limitations of existing methods, this paper proposes distribution network electrical equipment defect identification based on multi-modal image voiceprint data fusion and channel interleaving. First, image and voiceprint feature models are constructed using two-dimensional principal component analysis (2DPCA) and the Mel scale, respectively. Multi-modal feature fusion is achieved using an improved transformer model that integrates intra-domain self-attention units and an inter-domain cross-attention mechanism. Second, an image and voiceprint multi-channel interleaving model is applied. It combines channel adaptability and confidence to dynamically adjust weights and generates defect identification results using a weighting approach based on output probability information content. Finally, simulation results show that, under the dataset size of 3300 samples, the proposed algorithm achieves a 8.96–33.27% improvement in defect recognition accuracy compared with baseline algorithms, and maintains an accuracy of over 86.5% even under 20% random noise interference by using improved transformer and multi-channel interleaving mechanism, verifying its advantages in accuracy and noise robustness. Full article