最新刊期
- 摘要：α-Calcium sulfate hemihydrate (α-CSH) has problems such as excessive degradation, low mechanical strength and insufficient biological activity. There is also a risk of infection in clinical use, which limits its application in the field of bone repair. In order to solve the above problems, calcium sulfate dihydrate (CaSO4·2H2O) was used as raw material to prepare α-CSH by inorganic salt solution modification and autoclave dehydration. Then, ZnO/CSH composites with good mechanical properties, antibacterial properties and biocompatibility were prepared by compounding α-CSH with 2% nano-zinc oxide (ZnO). The results show that the composite has higher compressive strength (33.84 MPa), which is 46.24% higher than that of pure CSH. It has good antibacterial properties, and the antibacterial rates against Escherichia coli and Staphylococcus aureus reached 93.68% and 97.58%, respectively. It had better biological activity, and the proliferation activity of osteoblasts reached 110.52% on the 7th day. The composite material is expected to be used as a composite artificial bone in the field of infectious bone defect repair.关键词：α-calcium sulfate hemihydrate;nano-zinc oxide;compressive strength;antibacterial property;bioactivity;composite artificial bone16|13|0更新时间：2026-07-10
- 摘要：Using density functional theory (DFT) together with the self-consistent reaction field (SCRF) theory, the reaction mechanism of phenylalanine–magnesium (Phe→Mg2+) with hydroxyl radical (·OH) in aqueous solution was investigated at the dual-level theory of MN15/SMD/6-311++G(3df,2pd)//M06-2X/SMD/6-311+G(d,p). The reactions between Phe→Mg2+ and ·OH proceed through three pathways: hydrogen abstraction, addition, and electron transfer. Computational results show that, under implicit water solvent effect, the H-abstraction reactions by ·OH are all thermodynamically favorable processes, with activation energy barriers ranging from 41.2 to 52.0 kJ/mol. The addition reactions of ·OH with the carboxyl carbon are thermodynamically and kinetically unfavorable, while those with the benzene ring carbons are thermodynamically favorable, with barriers of 30.5-36.1 kJ/mol; the electron transfer process is thermodynamically and kinetically unfavorable. In the presence of explicit solvent water molecules through hydrogen bonding and coordination, the H-abstraction reactions are all thermodynamically favorable processes, with barriers decreasing to 15.7-25.1 kJ/mol, and the addition reactions of ·OH with the ring carbons become thermodynamically favorable, with barriers reduced to 14.1-22.2 kJ/mol. Moreover, the barrier for the electron transfer process decreases significantly from 249.0 to 117.6 kJ/mol. These results demonstrate that Phe→Mg2+ can rapidly scavenge hydroxyl radicals through both hydrogen abstraction and addition reactions in aqueous solution.关键词：magnesium-phenylalanine;hydroxyl radical;density functional theory (DFT);hydrogen bond;coordination bond;Gibbs free energy barrier18|14|0更新时间：2026-07-08
- 摘要：Dialect speech exhibits significant variations in pronunciation patterns, acoustic distributions, and linguistic structures, which poses challenges to end-to-end speech recognition models, especially under low-resource conditions. This paper proposes a Sichuan dialect speech recognition method based on multi-scale convolutional Conformer by optimizing both encoding and decoding stages. The model is composed of a multi-scale convolutional encoder and a hybrid decoder jointly modeled by connectionist temporal classification (CTC) and attention mechanism. In the encoding stage, a multi-scale convolutional structure is incorporated into the convolution module of Conformer to achieve parallel extraction of local acoustic features across different temporal scales, and a gated fusion mechanism is employed to adaptively integrate multi-scale representations, enhancing the modeling of complex dialect speech characteristics. In the decoding stage, a hybrid decoding architecture of CTC and bidirectional Transformer is adopted, integrating a specific N-gram language model trained on Sichuan dialect text, and a dynamic attention rescoring strategy is proposed to adaptively adjust the fusion weights between acoustic and language model scores based on decoding path statistics. Experimental results on a real Sichuan dialect speech dataset show that the proposed method reduces the character error rate by approximately 17%, which demonstrates the effectiveness of the proposed approach for Sichuan dialect speech recognition.关键词：speech recognition;dialect speech recognition;Conformer;end-to-end model10|6|0更新时间：2026-07-07
- 摘要：To address the limitations of traditional classroom behavior assessment, including low efficiency, strong subjectivity, and limited scalability, this study investigates student classroom behavior recognition based on video action analysis. A classroom behavior dataset was constructed from real teaching scenarios, containing three representative behaviors: attentive listening, looking down, and sleeping. Based on this dataset, a hierarchical spatiotemporal perception framework is proposed. A global spatiotemporal enhancement mechanism is introduced to improve long-range dependency modeling and hierarchical feature representation. In addition, a context-enhanced feature fusion strategy is employed to integrate local behavioral cues with global contextual information. To alleviate class imbalance, an adaptive class-balanced loss is designed to dynamically adjust the contributions of samples with different confidence levels. Experimental results demonstrate that the proposed method can accurately recognize student behaviors in complex classroom environments involving multiple targets and occlusions. Ablation studies further verify its effectiveness and robustness in recognizing behaviors with subtle motion variations and fine-grained differences.关键词：classroom behavior analysis;video action recognition;multi-scale temporal sampling;adaptive balance-aware loss function7|12|0更新时间：2026-07-03
- 摘要：To address the security issues caused by using partial hashes to verify file ownership in traditional data backup systems, the Proof of Ownership (PoW) mechanism has been introduced into data deduplication systems. However, existing PoW schemes still have certain limitations. Some schemes do not design ownership verification mechanisms based on complete files, and traditional schemes can hardly resist the increasing threat of quantum attacks. To address these issues, this paper proposes a proof of ownership verification scheme based on chained hashing and lattice-based zero-knowledge proof, namely HC_ZKP_PoW. In HC_ZKP_PoW, the complete file to be verified is divided into fixed-size data blocks, and hash values are iteratively computed block by block to generate a hash root that binds the file content and block order. Any tampering, deletion, or reordering of data blocks will change the hash root, and file integrity verification can be performed based on the hash root retained by the user. In addition, HC_ZKP_PoW adopts a lattice-based zero-knowledge proof mechanism for file ownership verification. Its security is based on the hardness of lattice problems, for which no known quantum polynomial-time algorithm can currently provide an efficient solution; therefore, the proposed scheme has post-quantum security advantages. Security analysis shows that HC_ZKP_PoW achieves favorable resistance to forgery and quantum attacks and can effectively protect data privacy. An attacker cannot pass the ownership verification without possessing the complete file, nor can the attacker obtain any private information related to the file from the verification interaction.关键词：data deduplication;proof of ownership;zero-knowledge proof20|36|0更新时间：2026-06-18
- 摘要：This study investigates the typical sandstone from the Jinjiang Landslide in the Baihetan Reservoir area of the Jinsha River. The deterioration effects of seismic activity and water-rock interaction on the mechanical properties of sandstone were simulated through wet-dry cycling and triaxial cyclic loading-unloading tests, respectively. The degradation patterns of dynamic and static parameters were analyzed, and the deterioration mechanisms were elucidated by integrating microstructural evolution characteristics. The results demonstrate that: with increasing test cycles, the damping ratio of sandstone gradually increases while the dynamic elastic modulus progressively decreases, with variations reaching 40.08% and 46.76% after 12 cycles, respectively. Static characteristic parameters-including triaxial compressive strength, cohesion, and internal friction angle-exhibit significant degradation, with maximum deterioration magnitudes of 37.30%-40.98%, 29.46%, and 19.48%, respectively, accompanied by a transition in failure mode from tensile to shear. The deterioration of both dynamic and static mechanical properties of sandstone is attributed to reservoir-induced seismicity and water-rock interaction, which cause cementation loss of mineral particles and the development and coalescence of pores and fractures. The anticipated findings are applicable to rock mass stability assessment in reservoir areas.关键词：Baihetan Reservoir;cyclic loading and unloading;dry-wet cycle;sequential effect;failure form;degradation mechanism616|27|0更新时间：2026-06-16
- 摘要：To address the challenges faced by existing Encrypted Traffic Analysis (ETA) methods, including strong dependence on labeled data, task-specific model design, and insufficient cross-scenario generalization capability, this paper proposes a pre-training and fine-tuning framework based on self-supervised contrastive learning. To better adapt to the temporal and protocol characteristics of traffic data, a MASK-based traffic augmentation method and a homologous replacement-based traffic augmentation method are designed to construct positive and negative sample pairs required for contrastive learning. Large-scale unlabeled traffic data are utilized for self-supervised pre-training, enabling the model to learn universal representations of encrypted traffic from structural characteristics and temporal behavioral patterns. High-performance cross-scenario transfer is then achieved through fine-tuning with only a small amount of labeled data. Experimental results demonstrate that the proposed approach outperforms existing methods across four downstream tasks, including identification of the emerging DoH (DNS over HTTPS) protocol, IoT (Internet of Things) attack behavior detection, anonymous network traffic identification, and malware identification. By leveraging pre-training, the proposed framework effectively reduces the demand for labeled data and provides an efficient and generalizable solution for encrypted traffic analysis.关键词：encrypted traffic analysis;self-supervised learning;contrastive learning;pre-training and fine-tuning;traffic augmentation32|48|0更新时间：2026-06-08
- 摘要：To address the unclear deformation mechanisms and insufficient displacement prediction accuracy of step-like landslides under the combined influence of rainfall and reservoir water level fluctuations, this study focuses on Baijiabao landslide on the right bank of the Xiangxi River in the Three Gorges Reservoir area. Through geological surveys, multi-source monitoring, and numerical simulations, the deformation evolution mechanisms of the landslide were systematically analyzed, and a displacement-oriented hybrid prediction model based on Variational Mode Decomposition (VMD), Improved Dung Beetle Optimization (IDBO), and Long Short-Term Memory (LSTM), was developed. The research results indicate that: 1) The mismatch between the reservoir water level decline rate and the seepage response rate of the low-permeability sliding mass leads to increased excess pore water pressure within the slip zone, which is the main cause of triggering step-like displacement; 2) The influence of rainfall on landslide deformation is relatively weak, mainly manifested as a superimposed effect on deformation at the rear edge of the landslide; 3) Reservoir water level fluctuations alter the seepage field in the slope, resulting in a staged variation characteristic of “enhancement-stabilization-weakening” in landslide stability; 4) The hybrid model (VMD-IDBO-LSTM) constructed based on displacement data of Baijiabao landslide demonstrates significantly better prediction accuracy than comparative models (VMD-LSTM and VMD-DBO-LSTM), with a root mean square error (RMSE) of 8.894 mm and a correlation coefficient (R²) of 0.983, effectively predicting step-like landslides in the reservoir area.关键词：step-like landslide;Baijiabao landslide;reservoir water level fluctuation;deformation mechanism;displacement prediction683|54|0更新时间：2026-06-05
- 摘要：As software systems grow in scale and complexity, traditional program verification faces severe challenges in efficiency and reliability. Although formal verification provides rigorous correctness guarantees, it often suffers from state-space explosion when handling large-scale programs. To address this limitation, this paper proposes two LLM-based verification enhancement methods: data-scale optimization and counterexample generation. The former analyzes the relationship between program semantics and data scale, reduces verification scale while preserving semantic logic, and uses data-flow slicing to ensure reliability. The latter constructs potential counterexamples with an LLM and validates them through testing. Furthermore, this paper builds a parallel verification framework integrating LLMs and traditional verification tools, executing the two methods and the baseline tool concurrently and adopting the earliest validated result. Experiments on the SV-COMP benchmark show that the proposed methods and framework improve task coverage and resource efficiency while maintaining high verification accuracy, providing an effective solution for intelligent program verification.关键词：program verification;formal methods;large language models14|45|0更新时间：2026-06-02
- 摘要：As user queries grow increasingly complex, existing Retrieval-Augmented Generation (RAG) models tend to treat questions as flattened token sequences during query processing, overlooking critical syntactic structure information and consequently degrading their ability to handle complex queries. To address this issue, this paper proposes a syntax-aware GraphRAG framework. The core innovation lies in incorporating dependency-based syntactic structures into the GraphRAG framework while employing Low-Rank Adaptation (LoRA) for efficient parameter adaptation. Specifically, prior linguistic features derived from dependency syntax trees are injected into both the retrieval and generation phases of the framework; meanwhile, LoRA is employed to achieve efficient parameter adaptation and fine-tuning of the LLaMA-3.1-8B-Instruct model. Experiments are conducted on two widely-used datasets, QALD-h and DBpedia QA, and the results demonstrate that our proposed framework outperforms baseline models across all key metrics. Ablation studies further confirm the performance gains brought by the incorporation of syntactic structures and parameter-efficient fine-tuning.关键词：retrieval-augmented generation;graph-based retrieva;syntax-aware;low-rank adaptation;parameter-efficient fine-tuning55|58|0更新时间：2026-06-02
- 摘要：Focusing on the stress and water occurrence environment of mudstone in the bank slope water-level-fluctuation zone of the Jurassic Shaximiao Formation in the Wanzhou section of the Three Gorges Reservoir area, coupled stress and wetting-drying cycle disintegration tests were conducted under different overburden stresses (0~0.6 MPa). The disintegration characteristics and energy evolution mechanism were analyzed. The results show that: 1) The coupled stress and wetting-drying cycles significantly accelerate the disintegration process of mudstone, with stress exerting an accelerating and regulating effect. 2) The accumulation rate and upper limit of surface energy increase with increasing overburden stress. The accumulation rate is highly consistent with the evolution of the disintegration rate, and the surface energy exerts a “promoting first and then inhibiting” effect on the disintegration process. 3) Under coupled action, the overburden stress exhibits stage-differentiated evolution characteristics: accelerated damage during the 1th-5th wetting-drying cycles, and promoted restructuring during 6th-12th cycles. The research results can provide support for the stability evaluation of sandstone-mudstone interbedded reservoir banks in the Three Gorges Reservoir area.关键词：mudstone in water-level-fluctuation zone;coupled stress and wetting-drying cycles;disintegration test;mesostructure;energy evolution579|37|0更新时间：2026-05-29
- 摘要：Efficient generation of reactive oxygen species (ROS) is critical for the degradation of organic pollutants in advanced oxidation processes. In this study, a Venturi cavitation device was employed to induce ROS generation via hydrodynamic cavitation. This work focused on the pathways of ROS formation and their effects on the degradation of organic pollutants. The results demonstrate that: 1) dissolved oxygen (O2) is a key source of ROS during hydrodynamic cavitation, and the main reactive oxygen species are 1O2 and O2; 2) Increasing the inlet pressure and solution temperature enhances the cavitation intensity, which is positively correlated with ROS generation; 3) The ROS generated in the device effectively degraded Rhodamine B (RhB), tetracycline (TC), methylene blue (MB), and methyl orange (MO), with removal efficiencies of 81%, 75%, 75%, and 65%, respectively, after 140 min of reaction; under optimal conditions, the degradation efficiency of RhB reached 100%. These findings reveal the intrinsic link between hydrodynamic cavitation and organic pollutant degradation, providing a theoretical basis and application prospects for pollutant removal via hydrodynamic cavitation.关键词：advanced oxidation process;hydrodynamic cavitation;organic pollutants;reactive oxygen species;formation pathway;numerical simulation26|48|0更新时间：2026-05-29
- 摘要：Intelligent analysis of economic benefits in sea ranching is essential for fine-grained aquaculture management. However, existing data-driven methods rely heavily on large labeled datasets and often suffer from underfitting in small-sample scenarios. To address this problem, this paper proposes two prompt-engineering-based intelligent analysis methods: a zero-shot economic benefit analysis method, named Prompt-Engineered Sea Pasture Economic Analysis (PE-SEA), and a sample-augmentation-based economic benefit analysis method, named Prompt-Augmented Pretraining for Sea Pasture Economic Analysis (PAP-SEA).Both methods use prompt engineering to guide large language models in analyzing marine ranching-related data and predicting future economic benefits. PE-SEA performs zero-shot inference based on GPT-4 and does not rely on additional training data; PAP-SEA uses ChatGPT for sample augmentation, while model training and inference are both completed locally. PE-SEA is more suitable for performance-prioritized scenarios, while PAP-SEA is more suitable for scenarios with higher requirements for data security and privacy protection. To facilitate practical application, this paper develops a real-time intelligent analysis tool called SeaPastureGPT for the economic benefits of marine ranching. Experiments are conducted on real monitoring datasets from three different sea areas and aquaculture species. The results show that both methods outperform baseline approaches. In particular, PE-SEA improves the F1-score by 70.67% compared withthe best-performing baseline method called KnowledgeGraph-BERT. Overall, our methods offer new insights and practical guidance for both academic research and industrial applications.关键词：marine pasture;intelligent economic benefit analysis;prompt engineering;large language model;pre-trained model;SeaPastureGPT42|52|0更新时间：2026-05-27
- 摘要：Monitoring the dynamic changes of cellular sulfur dioxide (SO2) is essential for elucidating its role as a gaseous signaling molecule in myocardial injury and apoptosis. Here, a dual-mode SO2 detection platform was constructed based on the synergistic operation of near-infrared ratiometric fluorescence imaging and a fluorescent probe functionalized field-effect transistor (FP-FET) sensor. The near-infrared fluorescent probe binds with SO2 both inside and outside the cells to achieve high-sensitivity fluorescence imaging, while the FP-FET sensor functionalized with the fluorescent probe converts the recognition events of SO2 released from the cells into quantifiable electrical signals, thereby enabling dynamic monitoring of cellular SO2 release. The FP-FET sensor exhibits excellent detection performance, with a detection limit as low as 0.05 pmol/L for SO2, along with outstanding specificity, repeatability, and stability. Using this platform, we dynamically monitored SO2 levels in H9c2 cardiomyocytes and in a myocardial ischemia-reperfusion injury (MIRI) model. The results demonstrate that the MIRI process is accompanied by the generation and release of SO2 from cells. The SO2 levels can serve not only as a quantitative indicator for evaluating the efficacy of ferroptosis inhibitors but also as a marker reflecting the extent of cellular injury and therapeutic prognosis during MIRI. This dual-mode strategy provides an effective tool for the simultaneous analysis of intracellular and extracellular signaling molecule dynamics under physiological and pathological conditions.关键词：field-effect transistor;fluorescence imaging;sulfur dioxide;myocardial ischemia-reperfusion injury;cell damage38|57|0更新时间：2026-05-22
- 摘要：Water-quality classification plays a key role in the maintenance and management of marine ranch, yet the heterogeneity of multi-source environmental factors makes it difficult for traditional methods to balance feature selection and interpretability. This study proposes a Multi-segment Sparse Mixture of experts Network (MSMN) that built a classification framework based on structured feature segments. The model embeds features and applies soft sparse filtering to highlight key segments while compressing low-contribution segments into background representations. A sparse mixture-of-experts module is then introduced, in which learnable routing is used to assign different segments to different experts, and multi-head attention is used to capture inter-segment interactions and the coupling relationships among environmental factors. Experiments on four Chinese coastal regions show that MSMN achieves higher accuracy and stronger minority-class recognition than conventional and deep models and maintains stable performance across regions and imbalance levels.关键词：fragment interaction;water quality classification of marine ranch;multi-expert system;self attention49|39|0更新时间：2026-05-22
- 摘要：The two-line sterile line RGD-7S was selected as the female parent and hybridized with the maintainer line Luohong 3B. After multiple generations of self-pollination purification, it was backcrossed with the sterile line Luohong 3A. In 2013, the indica Honglian type three-line sterile line Luohong 5A was selected and bred, which integrates disease resistance, fertility stability, and consistency of agronomic traits. In 2017, the research team successfully developed a new fifth-generation Honglian type hybrid rice variety, Luohong You 1564, using Luohong 5A as the female parent and the restorer line R1564 as the male parent. The variety displays an outstanding overall performance such as high yield, high-temperature tolerance, strong salt tolerance, resistance to rice blast and excellent grain quality. In the 2020—2021 national yield verification tests, the average yield in regional trials was 10.07 t/hm2 (671.4 kg/mu), an increase of 8.3% compared to the control variety Fengliangyou 4; the average yield in production trials was 9.74 t/hm2 (649.5 kg/mu), an increase of 9.1% compared to the control variety Fengliangyou 4. In 2025, during the super rice yield measurement and acceptance conducted by the Ministry of Agriculture and Rural Affairs, Luohong You 1564 achieved a yield of about 11.90 t/hm2 (793.1 kg/mu). In 2022, Luohong You 1564 was approved by the National Crop Variety Approval Committee (National Approved Rice 20220152). It was listed among the Leading Agricultural Varietiesin in Hubei Province for 2024—2025. In 2026, it was officially recognized as a super rice by the Ministry of Agriculture and Rural Affairs.关键词：hybrid rice;three-line sterile line;Honglian type;Luohong 5A;Luohong You 1564;breed selection;resistance to rice blast115|107|0更新时间：2026-05-21
- 摘要：To investigate the decision-making issues of battery manufacturers participating in cooperative recycling of power batteries, a closed-loop supply chain decision model for power batteries is constructed, comprising battery manufacturers, third-party recyclers, and cascade utilization enterprises. The study compares and analyzes the optimal supply chain decisions for battery manufacturers and third-party recyclers under cost-sharing and responsibility-sharing cooperative recycling models. Research findings indicate that participation of power battery manufacturers in cooperative recycling consistently improves battery recovery rates, boosts their own profits, and increases supply chain system profitability. Responsibility-sharing cooperation outperforms cost-sharing models, though cooperative recycling approaches do not influence the decision of cascade utilization providers. When adopting responsibility-sharing cooperative recycling, the manufacturers’ share ratio should be below a specific threshold(i.e.,0.670 4) to facilitate collaboration. Reasonable allocation of recycling responsibilities serves to maximize the profit outcomes for individual supply chain members while optimizing the overall supply chain system profits. Additionally, third-party recyclers can achieve a mutually beneficial outcome for all companies for the supply chain by appropriately lowering recycling prices.关键词：power battery;collaborative recycling;cascade utilization26|41|0更新时间：2026-05-21
- 摘要：Intelligent technologies can accelerate collaborative knowledge construction, yet it remains unclear which specific links they act on and what level of automation is suitable for which educational scenarios. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), this study conducts a systematic review of 29 relevant literatures from 2014 to 2024. It is found that the period from 2014 to 2020 was the initial exploratory stage, followed by a period of rapid growth from 2021 to 2024, which is highly consistent with the advancement of global education digitalization strategies and technological breakthroughs such as generative artificial intelligence. Specific research findings are as follows: 1) Current research is characterized by diverse participants, integrated methods, and differentiated contexts, dominated by China and the US, centered on students, mainly focused on higher education and online learning, with mixed methods as the main research method; 2) Intelligent technology empowers collaborative knowledge construction mainly through diverse technological collaboration, automated support, and multimodal data fusion, with machine learning, deep learning, and natural language processing as core technologies, and highly automated student-centered models as the dominant approach; 3) The impact of intelligent technology on collaborative knowledge construction concentrates on core processes such as knowledge integration and sharing, relying primarily on automated analysis, with outcomes focusing on skill improvement. Research adopts a cross-perspective dominant view, yet imbalances remain, including an overemphasis on cognition over affect and insufficient support for peer interaction. Future research should enhance the automated adaptability of intelligent technologies across educational stages, clarify their functional mechanisms in knowledge construction, and extend applications to interdisciplinary and blended learning contexts, thereby advancing intelligent technologies from instrumental support toward ecological integration.关键词：intelligent technologies;collaborative knowledge construction;systematic literature review130|138|0更新时间：2026-04-17
- 摘要：With the rapid development of generative AI, large language models are increasingly used in language education. This study examines the ethical boundary between “optimization” and “overstepping” in L2 writing by investigating the linguistic appropriateness of the reasoning model DeepSeek-R1 in secondary-school English writing. Using a mixed design that combines paired student texts and a teacher survey, we apply paired comparisons and regression analyses across language quality, editing behavior, and teacher perceptions. Results show that DeepSeek-R1 increases lexical diversity and gradually shifts lexical sophistication toward mid-to-high CEFR levels, while maintaining relatively high semantic preservation despite substantial edits. Revisions are mainly replacement-based paraphrasing, with no large loss of content words or systematic syntactic or discourse restructuring, suggesting a language-polishing role rather than cognitive substitution. These findings systematically present a multi-level, replicable quantitative evidence base characterizing LLM-mediated writing feedback in secondary-school L2 contexts, providing an evidence-based framework for classroom governance of LLM writing assistance.关键词：DeepSeek;generative AI;L2 writing;appropriateness249|214|0更新时间：2026-04-10
- 摘要：Accurately detecting grid synchronization signals using a phase-locked loop is a crucial prerequisite for ensuring the safe and stable operation of power electronic equipment in power systems. However, in the context of the “double high” power system, synchronous detection of the power grid faces complex disturbances such as harmonic distortion and three-phase imbalance. To this end, based on the Delayed Signal Cancellation (DSC) filtering principle, this paper proposes an enhanced cascaded approximation structured PLL structure. Initially, by cascading multiple DSC operators with distinct delay parameters, a Cascade Delayed Signal Cancellation (CDSC) pre-filtering structured PLL is constructed. This configuration enables precise detection of fundamental positive-sequence components under harmonic distortions and unbalanced grid conditions, while effectively preventing the interference of disturbance components on the PLL path. Secondly, incorporating the grid frequency obtained from the PLL output, a frequency feedback mechanism is introduced where delay parameters of individual DSC operators are dynamically calculated through sampling frequency analysis, achieving adaptive dynamic adjustment of the pre-filtering system under frequency fluctuations. Finally, to address fractional delays in the feedback loop, an approximation computation method cascaded within DSC operators is proposed, which enhances the precision of positive-sequence fundamental component tracking and harmonic elimination in the CDSC filtering structure. Experimental results demonstrate that the proposed enhanced cascaded approximation structure PLL maintains stable grid synchronization signal detection under harmonic distortions, frequency fluctuations, and three-phase unbalance conditions, with the total harmonic distortion of detected fundamental positive-sequence voltage remaining below 0.6% and frequency measurement errors constrained within 0.05 Hz and phase detection error within 0.01 rad.关键词：phase-locked loop;delay cancellation;frequency adaptation;harmonic distortion50|181|0更新时间：2026-04-08
- Tel：027-68756952 Email：whdz@whu.edu.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
- Total Visits：1427450 Visits Today：511
Bulk Citation
0