• Volume 41,Issue 20,2025 Table of Contents
    Select All
    Display Type: |
    • >Discipline Development
    • Research on development of comprehensive practical training teaching material for industrial internet major based on “1+X” certificate system

      2025, 41(20):1-3.

      Abstract (156) HTML (0) PDF 1.29 M (516) Comment (0) Favorites

      Abstract:To address issues such as the misalignment between comprehensive training textbooks for the Industrial Internet major and the requirements of the “1+X” certificate, as well as delays in technology iteration, this article proposes a “industry demand-teaching content-certification” ternary integration model for training textbook development. By aligning with the competency framework (primary, intermediate, and advanced) of the Industrial Internet Implementation and Operation & Maintenance vocational skill level standards, a hierarchical and progressive modular training textbook system is constructed. In addition, the use of a loose-leaf textbook structure combined with digital resource packages enables dynamic content updates, providing a replicable paradigm for the development of training textbooks in vocational education.

    • Research on construction method of interdisciplinary educational resource database for vocational college with GenAI technology

      2025, 41(20):4-6.

      Abstract (162) HTML (0) PDF 1.24 M (470) Comment (0) Favorites

      Abstract:In response to issues such as the fragmentation of interdisciplinary resources and low efficiency in industry-education collaboration in the digital transformation of vocational education, this paper proposes a method for constructing an interdisciplinary educational resource repository for vocational colleges based on GenAI technology. By integrating dynamic knowledge graphs with multimodal generation models, it constructs a tripartite technical framework of “intelligent resource production-precise teaching adaptation-continuous ecological evolution”. Key technologies such as domain-adaptive pre-training, graph diffusion association inference, and knowledge-guided adversarial training are employed to achieve automated generation and dynamic optimization of interdisciplinary educational resources. Experiments show that this method increases the resource update frequency by 598%, boosts learners? skill achievement rate by an average of 29.0%, and improves job match accuracy to 81.6%. Additionally, it constructs a trustworthy shared ecosystem through blockchain smart contracts and differential privacy protection. Practical validation demonstrates the effectiveness of this solution in promoting deep integration of industry and education and cultivating compound technical and skilled talents, providing a replicable technical paradigm for the digital transformation of vocational education.

    • Evaluation of effect of nursing professional curriculum teaching reform based on knowledge graph

      2025, 41(20):7-9.

      Abstract (180) HTML (0) PDF 1.26 M (482) Comment (0) Favorites

      Abstract:The traditional evaluation methods for teaching reform often focus on quantitative indicators such as exam scores and employment rates, and fail to fully incorporate qualitative dimensions such as students? comprehensive quality and innovation ability, resulting in significant deviations in the evaluation results. Therefore, this article proposes a method for evaluating the effectiveness of nursing course teaching reform based on knowledge graph. By constructing an evaluation index system based on knowledge graph and integrating the weights of various indicators, a comprehensive evaluation of the effectiveness of teaching reform can be achieved. The results of the case analysis show that this method can more accurately reflect the actual effectiveness of nursing professional course teaching reform and provide scientific basis for the promotion of teaching reform.

    • Typical case analysis of Excel IF function application in “information technology” course

      2025, 41(20):10-12.

      Abstract (186) HTML (0) PDF 1.44 M (450) Comment (0) Favorites

      Abstract:In the undergraduate “Information Technology” course, Excel commonly used functions are the key teaching content, among which the IF function is a core knowledge point that students must master. Using the “Student Performance” data table as the object, it is required to use the IF function to grade the average score. The specific criteria are as follows: if the average score is greater than or equal to 90, it is judged as “excellent”; A score greater than or equal to 80 is considered “good”. A score greater than or equal to 70 is considered “satisfactory”. A score greater than or equal to 60 is considered “passing”, If the average score is less than 60, it is judged as? failing?. By logical analysis of the conditions of the problem and combining the basic principles of IF functions, two formula writing processes for multi-layer nested IF functions ave designed. This case can visually demonstrate the application methods and design ideas of multi condition and multi-layer nesting, which helps students to deeply understand and master the application skills of IF functions in practical problems, thereby improving learning effectiveness.

    • Exploring path of deep integration between artificial intelligence and higher education

      2025, 41(20):13-15.

      Abstract (193) HTML (0) PDF 1.15 M (525) Comment (0) Favorites

      Abstract:The rapid development of artificial intelligence technology is profoundly reshaping the ecology of higher education. By analyzing the problems existing in the application of artificial intelligence in higher education, three integration paths are proposed. (1) Refactoring the teaching system. Relying on intelligent driven precision teaching mode, virtual teaching and research room collaboration, and multi-dimensional intelligent evaluation, we promote teaching from standardization to personalization.(2) Promoting the transformation of the roles of teachers and students. Constructing a dual helix competency model of “subject technology” for teachers, stimulating students subjectivity ? , and realizing the transformation of teaching methods from one-way imparting to symbiotic development.(3) Management of innovative education. Relying on the campus intelligent center, blockchain credit bank, and dynamic quality monitoring loop, we aim to achieve a leap in educational governance from empirical decision-making to data governance. Through systematic transformation, it is expected to break through the boundaries of traditional education and form a new educational paradigm of human-machine collaboration, providing theoretical reference and practical guidance for the digital transformation of higher education.

    • Realistic dilemma and breakthrough path of digital transformation in vocational college

      2025, 41(20):16-18.

      Abstract (184) HTML (0) PDF 1.16 M (513) Comment (0) Favorites

      Abstract:In the era of digital economy, the digital transformation of vocational colleges has become a key path to promote the high-quality development of vocational education. Based on literature analysis, case studies, and questionnaire surveys, the main challenges faced in the current transformation are systematically analyzed: insufficient supply and lagging updates of teaching resources. Teachers have weak digital literacy and low degree of technological integration. The management system is fragmented, and data efficiency is limited. Unequal allocation of resources and prominent differences in regional and intercollegiate development. Insufficient internal drive for students? digital learning. Weak network security protection. The teaching evaluation system is not perfect. A systematic breakthrough path is proposed to address the above issues: government school enterprise collaboration to jointly build a dynamic resource platform, hierarchical improvement of teachers? digital literacy, promotion of data-driven management innovation, optimization of resource balance allocation, innovation of student learning incentive strategies, strengthening network security guarantees, and improvement of multi-dimensional teaching evaluation system. Intended to provide theoretical support and practical reference for vocational colleges to break through the bottleneck of transformation and enhance governance efficiency.

    • Access control method for teaching and examination platform based on RBAC model

      2025, 41(20):19-21.

      Abstract (223) HTML (0) PDF 1.45 M (550) Comment (0) Favorites

      Abstract:To ensure the security of online teaching and examination platforms and prevent unauthorized access, a role-based access control (RBAC) method has been designed and implemented. Firstly, a teaching and examination platform based on SpringBoot and Layui technology stack was constructed, covering core functions such as question generation, paper generation, and online exams. On this basis, a three-layer database model of user role permission association was designed, and a flexible and configurable access control module was implemented using SpringBoot Interceptor technology. This module can perform pre permission verification before the user request reaches the business controller, effectively preventing unauthorized access. The system testing results show that this method can accurately control functional access permissions according to different roles (administrators, teachers, students), significantly improving the security and manageability of the platform.

    • Practical research on humanoid robot in “computer network” course

      2025, 41(20):22-24.

      Abstract (143) HTML (0) PDF 1.18 M (484) Comment (0) Favorites

      Abstract:As an important component of information technology education, the teaching methods of the “computer network” course are constantly being innovated, aiming to enhance students? practical abilities and innovative thinking. Introducing humanoid robots as practical teaching tools not only enhances the interactivity and interest of the classroom but also provides students with a network environment close to reality, promoting their in-depth understanding and application of core knowledge such as communication principles, data transmission mechanisms, and network security strategies. Through practical tasks that combine robot programming with network control, students can systematically cultivate logical thinking, engineering practice and problem-solving abilities in hands-on operations. The integrated application of humanoid robots in computer network teaching provides a new path for the cultivation of innovative information technology talents and has positive significance for promoting the reform of course teaching and the overall quality improvement.

    • Application of multimodal large models in vocational education training scenarios

      2025, 41(20):25-27.

      Abstract (199) HTML (0) PDF 1.19 M (536) Comment (0) Favorites

      Abstract:With the rapid development of artificial intelligence technology, Multimodal Large Models (MLLMs) demonstrate powerful cognitive and generative capabilities in processing multi-source information such as speech, images, text, and video. The demand for comprehensive perception, real-time interaction, and personalized guidance in vocational education training scenarios is highly compatible with the advantages of multimodal large models. This article combines the characteristics of vocational education to explore the application mode of multimodal large models in practical teaching, including task analysis, skill demonstration, process evaluation, and intelligent feedback. It also takes the manufacturing CNC machining and information technology virtual simulation training as a case study for analysis. Researchshows that MLLMs can not only improve practical training efficiency and teaching quality, but also optimize students? learning paths and achieve personalized teaching.

    • Research on construction of teaching evaluation system for computer major in private colleges under background of industry-education integration based on OBE-CDIO

      2025, 41(20):28-30.

      Abstract (190) HTML (0) PDF 1.17 M (433) Comment (0) Favorites

      Abstract:Under the background of new engineering construction, engineering education professional accreditation and industry-education integration, in response to the current problems of the teaching evaluation system for computer science in private colleges and universities, a teaching evaluation system that adapts to market demands and student development is onstructed by integrating the OBE (Outcome-Based Education) and CDIO (Conceive, Design, Implement, Operate) concepts. This system is oriented towards clear learning outcomes, covering multi-dimensional evaluation contents such as knowledge acquisition, ability attainment, and quality cultivation. It adopts diversified evaluation methods including process evaluation, value-added evaluation, and comprehensive evaluation, clarifies the responsibilities of evaluation subjects such as teachers, students, and enterprise mentors, and establishes an effective feedback and improvement mechanism, providing strong support for the cultivation of computer science talents in private colleges and universities.

    • Research on construction of ideological and political education in course of “foundation of big data technology application” under pattern of “big ideological and political education”

      2025, 41(20):31-33.

      Abstract (222) HTML (0) PDF 1.16 M (543) Comment (0) Favorites

      Abstract:Starting from the course of “foundation of big data technology application”, this article argues the necessity of carrying out ideological and political education in finance and economics majors (such as big data and accounting). Specific measures have been proposed around the implementation path of ideological and political education in the curriculum, including condensing ideological and political elements, improving teaching methods, enriching teaching tools, and optimizing teaching processes, with the aim of promoting the organic integration of “professional education” and “ideological and political education”. The case study of “personal Income tax Python program design” further demonstrates the combination of knowledge points and ideological and political elements, aiming to enhance the ideological and moral literacy of students majoring in finance, economics, and commerce, implement the fundamental task of “foster virtue through education” in universities, and integrate core values into the entire process of education and teaching.

    • Research on personalized teaching mode of information technology in higher vocational education driven by big data

      2025, 41(20):34-36.

      Abstract (147) HTML (0) PDF 1.18 M (541) Comment (0) Favorites

      Abstract:This article takes the Information Technology program in higher vocational education as a research subject to explore the construction logic and implementation pathways of a personalized teaching model driven by big data. By analyzing common issues in existing IT teaching practices and leveraging the collection, in-depth analysis, and learning prediction capabilities of big data, a personalized teaching framework based on learning behavior data, knowledge graphs, and intelligent recommendation algorithms was designed. The research demonstrates that this model effectively enhances teaching adaptability and precision, while fostering students? learning interest and professional competencies.

    • Nurturing young: knowledge mapping analysis of moral education in kindergarten from perspective of “Fostering Virtue through Education”(2015—2025)

      2025, 41(20):37-40.

      Abstract (184) HTML (0) PDF 1.52 M (531) Comment (0) Favorites

      Abstract:To systematically analyze the research progress and trends in the field of ideological and political education in kindergartens in China, this study uses literature from the China National Knowledge Infrastructure (CNKI) database published between 2015 and 2025. The CiteSpace tool was employed for bibliometric and knowledge mapping visualization analysis. The results show that:(1) Research in this field started relatively late but has developed rapidly in recent years, demonstrating a clear policy-driven feature, while collaboration networks among research institutions and authors remain to be strengthened.(2) Research hotspots focus on “curriculum-based ideological and political education,”“environment creation,” and the integration practices with the “five domains” of early childhood education, with the knowledge structure gradually shifting from localized exploration to systematic construction.(3) Research methods exhibit a trend of diversified integration, and the role of technological empowerment has become increasingly prominent.(4) Future studies should focus on exploring regionally characteristic paths, building collaborative mechanisms among families, kindergartens, and communities, and localizing international experiences.

    • Exploration of teaching “Java programming” course assisted by large model

      2025, 41(20):41-43.

      Abstract (139) HTML (0) PDF 1.24 M (481) Comment (0) Favorites

      Abstract:The rapid development of big model technology is driving continuous changes in higher education, promoting the deep integration of intelligent technology and higher education, and gradually forming a new model of intelligent teaching. Aiming at the problems of fixed teaching content, single teaching mode, and lack of personalized training in traditional “Java programming” course, a “Java programming” teaching system based on large model assistance is proposed. By utilizing large-scale modeling platforms such as Deepseek, practical problems in aerospace and other fields are introduced into Java teaching content, implementing the interactive teaching concept of deep integration between teachers, students, and large-scale models, thereby achieving personalized teaching and comprehensively enhancing students? system and innovation abilities.

    • Research on model of information technology talent training based on interdisciplinary integration under background of rural revitalization

      2025, 41(20):44-46.

      Abstract (238) HTML (0) PDF 1.18 M (602) Comment (0) Favorites

      Abstract:Under the strategic background of rural revitalization, there is an increasingly strong demand for talent cultivation, and it has become particularly urgent for universities to cultivate interdisciplinary and integrated information technology talents. This article aims to promote the integration of information technology with other disciplines, meet the demand for high-quality information technology talents in the agricultural industry, and provide reference for universities to build information technology talent training models through measures such as constructing interdisciplinary curriculum systems, establishing diversified innovative teaching models, and reforming diversified talent training systems.

    • Research on interactive teaching method of computer in applicationoriented university based on AIGC technology

      2025, 41(20):47-49.

      Abstract (150) HTML (0) PDF 1.36 M (568) Comment (0) Favorites

      Abstract:In the teaching reform of computer science in application-oriented universities, the traditional simple AI interactive teaching has the problem of content templating, resulting in a relatively low accuracy rate of teaching interaction (ACC). To this end, the article proposes a personalized interactive teaching method based on AIGC. Firstly, structured data is analyzed through a knowledge graph to obtain personalized teaching resources that are suitable for students? cognitive levels. Secondly, AIGC technology is utilized to precisely match resources with students? demands and dynamically generate interactive teaching content. Finally, the optimal interactive information is filtered and output based on the preset learning strategy. The experimental results show that under different concurrent student numbers, the ACC value of this method remains stable at over 90%, verifying its significant advantages in improving the accuracy and adaptability of interactive teaching.

    • Research on big data course teaching based on Docker

      2025, 41(20):50-52.

      Abstract (141) HTML (0) PDF 1.29 M (415) Comment (0) Favorites

      Abstract:In order to keep up with the trend of the times, universities are vigorously building majors related to big data. However, both teachers and students often face many difficulties when building a big data training environment. To solve this problem, Docker container technology is adopted to build an easy to learn and convenient training environment. This environment not only effectively solves the problem of inconvenient transfer of training environment, but also simplifies the process of environment construction, making teachers? teaching more efficient and students? skill levels significantly improved during training. Based on this practical training environment, students successfully won the gold medal in the 2024 World Vocational College Skills Competition. Finally, the article provides a detailed introduction to the construction process and operational steps of the training environment, providing replicable practical references for its smooth setup.

    • Research on cultivation path of teaching literacy for information majors in higher vocational education under background of digital transformation

      2025, 41(20):53-55.

      Abstract (142) HTML (0) PDF 1.26 M (505) Comment (0) Favorites

      Abstract:Against the backdrop of digital transformation in vocational education, information technology teachers in higher vocational education are facing dual challenges of accelerated technological iteration and teaching mode changes, which have raised higher requirements for teachers? teaching literacy. Based on the teaching literacy characteristics of information technology teachers and the demand for industry education integration in higher vocational education, the TPACK-IK model is established by introducing industry practice knowledge dimensions on the basis of TPACK theory. Through the linkage of practical empowerment provided by enterprises, institutional guarantees established by schools, and the enhancement of teachers? iterative abilities, a collaborative education ecosystem is formed, and a path for improving teachers? teaching literacy is constructed to achieve closed-loop improvement of teaching literacy.

    • Impact of AI technology on education in Anhui Province under background of leading country in education

      2025, 41(20):56-58.

      Abstract (173) HTML (0) PDF 1.22 M (451) Comment (0) Favorites

      Abstract:Based on the rapid development of AI technology in the field of education, this study comprehensively applies research methods such as text analysis, questionnaire surveys, interviews, and field research to deeply explore the impact of AI technology on education in Anhui Province. From the perspectives of four different roles: students, teachers, educators, and parents, this study analyzes their willingness to support the application of AI educational technology, with a focus on potential issues and influencing factors in current AI educational technology applications. On this basis, the application prospects of AI technology in the education field in Anhui Province were predicted, and practical and feasible development strategies for students, teachers, and society were proposed based on the actual situation in Anhui Province, providing important reference for the sustainable development of the education field under the background of big data and AI.

    • >Technology Frontier
    • Real-time early warning method of intelligent kitchen fire risk based on multi-modal perception

      2025, 41(20):59-61.

      Abstract (180) HTML (0) PDF 1.26 M (442) Comment (0) Favorites

      Abstract:To improve the reliability of intelligent kitchen systems, this article proposes a real-time fire risk warning method based on multi-modal perception. Using similarity indicators and recurrent neural networks to fuse multi-modal perception data and construct a fire risk warning indicator system. On this basis, a real-time warning model is established by combining the principle of information diffusion and fuzzy mathematical methods. The kitchen fire risk is divided into five levels to achieve dynamic assessment and warning of its risk status. The experimental results show that the warning results of this method are completely consistent with the actual occurrence of fires, with a low and steady increase in false alarm rate, and a false alarm rate always below 5%. It has high accuracy and reliability in fire risk warning.

    • Small target detection algorithm for UAV aerial photography based on improved XBE-YOLO

      2025, 41(20):62-64.

      Abstract (200) HTML (0) PDF 1.26 M (576) Comment (0) Favorites

      Abstract:In response to the problems of complex background, prominent false and missed detections, and low detection accuracy in UAV aerial image target detection, this article proposes a UAV aerial small target detection algorithm based on improved XBE-YOLO. Firstly, by abandoning the large target detection head and adding a small target detection head, the models feature perception ability for small targets is enhanced, thereby improving its detection performance. Secondly, a Bidirectional Feature Pyramid Network (BiFPN) is introduced to replace the Concat module in the original model, in order to enhance the fusion of multi-scale contextual information and further improve the detection accuracy of small targets. In addition, replacing the PSA module with an EMA lightweight attention mechanism effectively reduces the number of network parameters and computational complexity while maintaining detection accuracy. The experimental results on the VisDrone-2019 dataset show that the average detection accuracy of the improved model reaches 34%, which is 1.7% higher than the original YOLOv10n, verifying the effectiveness of the proposed improvement strategy.

    • Research on visualization technology of hydrological and water resources data base plate based on GIS

      2025, 41(20):65-67.

      Abstract (170) HTML (0) PDF 1.51 M (643) Comment (0) Favorites

      Abstract:With the rapid development of geographic information system (GIS) technology, its application in hydrological and water resources management is becoming increasingly widespread. The article systematically studied the GIS based visualization technology for hydrological and water resources data, focusing on key methods such as data collection, preprocessing, spatial analysis, and visualization rendering. Based on the ArcGIS platform and Python tools, the visualization expression of hydrological elements at multiple scales has been achieved, and the practical application effect of this technology in water resources management has been analyzed through practical cases. Research has shown that GIS based hydrological and water resources data visualization technology can significantly enhance the intuitive understanding of data and improve the scientificity of decisionmaking processes.

    • Classification method for human-computer interaction actions in fixed scenarios: self training lightweight scheme based on MoveNet and Naive Bayes

      2025, 41(20):68-70.

      Abstract (115) HTML (0) PDF 1.62 M (534) Comment (0) Favorites

      Abstract:The article proposes a human skeletal action classification method that is oriented towards lowperformance processing environments and supports self-training. It extracts key point information of the human body based on the lightweight MoveNet model and classifies the skeleton sequence using the Naive Bayes algorithm. The experimental results show that, under the condition that the user and the device maintain a fixed relative position, this method only needs to collect 15 to 20 seconds of user-customized data to achieve lightweight classification with high accuracy (about 95%). However, this method still has the risk of misrecognition for actions with subtle posture differences, and its generalization ability for unseen data is limited.

    • Research on code generation optimization driven by large model

      2025, 41(20):71-73.

      Abstract (170) HTML (0) PDF 1.29 M (505) Comment (0) Favorites

      Abstract:To enhance the structural integrity and semantic fidelity of large models in code generation tasks, this article constructs a multi-level optimization framework and systematically analyzes its mechanism from dimensions such as training data, algorithm structure, model architecture, and generation control. The experimental results show that the optimized model significantly outperforms existing mainstream baseline models in terms of functional accuracy, syntax consistency, and engineering maintainability, demonstrating stronger industrial adaptability and generation efficiency.

    • Application research of deep learning-based multimodal data fusion method in short video sentiment analysis

      2025, 41(20):74-76.

      Abstract (259) HTML (0) PDF 1.20 M (508) Comment (0) Favorites

      Abstract:With the rapid development of short video platforms, sentiment analysis of short video content becomes the key to understanding user behavior and improving content recommendation effectiveness. This article proposes a multimodal data fusion method based on deep learning, which combines visual, auditory, and textual information of videos for short video sentiment analysis. Comparative experimentsshows that this method outperforms traditional single modal analysis methods in both emotion classification accuracy and F1 score, providing a new approach for short video emotion analysis.

    • Automated detection of pests and diseases in multimodal agricultural crop based on knowledge graph

      2025, 41(20):77-79.

      Abstract (167) HTML (0) PDF 1.38 M (509) Comment (0) Favorites

      Abstract:The spread of pests and diseases poses a serious threat to food security and agricultural product quality. To achieve effective prevention and control of crop diseases and pests, this article studies a multimodal automated detection method based on knowledge graph. Firstly, entities, attributes, and relationships are extracted from multi-source heterogeneous data. Secondly, a knowledge graph is constructed, and based on this, a breadth first search algorithm is used to infer multimodal types of crop pests and diseases, thereby achieving automatic detection of pests and diseases. The experimental results show that as the sample size increases, the F1 value of this method remains at a high level, indicating its high accuracy and reliability in crop pest and disease detection.

    • Path to improving accuracy of electronic information image recognition based on deep learning

      2025, 41(20):80-82.

      Abstract (133) HTML (0) PDF 1.20 M (486) Comment (0) Favorites

      Abstract:With the rapid development of electronic information technology, image recognition has been widely applied in many fields. Deep learning has brought new opportunities for the development of electronic information image recognition, but recognition accuracy in this field still faces many challenges. This article delves into the path of improving the accuracy of electronic information image recognition based on deep learning, including data augmentation, model optimization, introduction of attention mechanisms, and multimodal fusion. Through experimental verification, these paths can effectively improve the recognition accuracy of electronic information images and provide more reliable technical support for related applications.

    • Method for fault detection of modern building electrical equipment based on deep learning algorithm

      2025, 41(20):83-85.

      Abstract (134) HTML (0) PDF 1.44 M (492) Comment (0) Favorites

      Abstract:The margin between the actual parameters of the equipment and the rated parameters may cause the fault characteristics to shift, thereby resulting in a significant deviation between the fault detection results based on feature analysis and the actual state. To this end, the article proposes a modern fault detection method for building electrical equipment based on deep learning algorithms. Firstly, the original fault data is preprocessed such as filling in missing values, and then the complete fault features are extracted by using convolution operations. In the stage of feature deep learning, an adaptive penalty factor is introduced, and the nonlinear distribution function of the fault characteristics of building electrical equipment is used as its constraint condition. The precise discrimination of the fault state is achieved by means of the convergence margin of the penalty term. The experimental results show that this method can effectively converge the influence brought by the parameter margin of the equipment, and the fault detection results have a high degree of agreement with the actual state.

    • Research on application of mechanical and electrical systems in intelligent manufacturing

      2025, 41(20):86-88.

      Abstract (125) HTML (0) PDF 1.20 M (511) Comment (0) Favorites

      Abstract:Firstly, the concept and characteristics of intelligent manufacturing are elaborated, emphasizing its key position in the development of modern industry. Secondly, the constituent elements of mechanical and electrical systems are analyzed, and their specific application scenarios in intelligent manufacturing are explored in depth, including automated production lines, intelligent robots, intelligent warehousing and logistics, aiming to promote the wider and more effective application of mechanical and electrical systems in intelligent manufacturing, and promote the transformation and upgrading of the manufacturing industry towards intelligence and efficiency.

    • Research on ultrasonic defect identification of new energy wind turbine blad based on machine learning

      2025, 41(20):89-91.

      Abstract (161) HTML (0) PDF 1.37 M (475) Comment (0) Favorites

      Abstract:As a core component of new energy equipment, the traditional detection methods for wind turbine blades have relatively low identification efficiency. To enhance the efficiency and accuracy of defect recognition, this article proposes a defect recognition method for ultrasonic testing based on machine learning. This method employs convolutional neural networks to automatically extract and classify typical defects in the composite material structure of blades, achieving high-precision intelligent recognition. The experimental results show that the recognition accuracy of this method reaches over 94%, and it has good robustness and engineering deployment capabilities, making it suitable for intelligent application scenarios in wind power operation and maintenance.

    • Improved lightweight strip steel surface defect detection based on YOLOv5

      2025, 41(20):92-94.

      Abstract (150) HTML (0) PDF 1.53 M (546) Comment (0) Favorites

      Abstract:An improved lightweight detection algorithm based on YOLOv5 is proposed to address the problems of high computational complexity and difficult deployment of models in surface defect detection of strip steel. Firstly, EfficientNetV2 is used to replace the original backbone network CSPDarknet-53, significantly reducing the model complexity. Secondly, the Ghost module is introduced into the neck network to construct a lightweight feature extraction structure, further reducing computational resource consumption. Finally, the content aware feature recombination operator is utilized to enhance the multi-scale feature fusion capability. The experimental results on the NEU-DET dataset show that the improved model has only 4.1 MB of parameters and 5.0 GB of floating-point operations, which are reduced by 40.9% and 68.2% respectively compared to the original YOLOv5. In terms of detection performance, the average accuracy has improved by 2.1%, and the detection speed has accelerated by 10.5%. This algorithm achieves model lightweighting while ensuring detection accuracy, providing an effective solution for real-time defect detection in industrial sites.

    • Hybrid document based on rule and deep learning directory location and information extraction technology

      2025, 41(20):95-97.

      Abstract (117) HTML (0) PDF 1.26 M (459) Comment (0) Favorites

      Abstract:Aiming at the problem of insufficient flexibility of traditional rule-based methods when dealing with documents of diverse formats and complex structures, this article proposes a hybrid document directory location and information extraction technology that combines rule-based methods with deep learning. This method fully exploits the deterministic nature of the rule-based approach and the adaptive advantages of deep learning. Firstly, it conducts preliminary screening and preprocessing of documents through rules, and then uses deep learning models to complete the recognition of complex patterns and semantic understanding. The experimental results show that this method significantly improves the accuracy of document directory location and the completeness of information extraction, providing a more reliable and efficient solution for the structured processing of document information and knowledge mining.

    • Research on 3D point cloud processing algorithm for high-precision online detection of steel pipe dimensions

      2025, 41(20):98-100.

      Abstract (156) HTML (0) PDF 1.21 M (460) Comment (0) Favorites

      Abstract:This article proposes a high-precision online detection algorithm framework based on 3D point cloud processing to meet the accuracy and efficiency requirements of dimension detection in the production process of steel pipes. By integrating multi view laser scanning and deep learning techniques, combined with dynamic registration and feature optimization strategies, real-time measurement of key parameters such as steel pipe outer diameter, wall thickness, and ellipticity is achieved. The experimental results show that the algorithm has an error of ≤±0.05 mm in diameter measurement and a wall thickness detection accuracy of ±0.03 mm, which is more than 40% higher than traditional methods and meets the quality control requirements of Industry 4.0 production lines.

    • Research on mangrove semantic segmentation network based on improved UVSAM

      2025, 41(20):101-102.

      Abstract (154) HTML (0) PDF 1.19 M (520) Comment (0) Favorites

      Abstract:The UVSAM model achieves fine-grained image segmentation through a hybrid prompt mechanism. However, the existing methods have deficiencies in feature fusion, resulting in limited segmentation accuracy. The article proposes an improved UVSAM model. By designing a new hybrid prompt generation module, the multi-scale feature fusion ability is enhanced, and a mask adaptive weighted fusion strategy is introduced to improve the robustness of the model. The experimental results on the FJJMangrove dataset show that the IoU index of the improved model reaches 85.42%, the F1 score is 79.80%, and the Precision is 85.41%. All performances are significantly better than the original UVSAM and other comparison models, verifying the effectiveness of this method.

    • Application and practice of AI algorithms in power system fault diagnosis and early warning

      2025, 41(20):103-105.

      Abstract (180) HTML (0) PDF 1.20 M (513) Comment (0) Favorites

      Abstract:With the expansion of the power system scale and the increase in the proportion of new energy, traditional fault diagnosis methods face challenges such as high data complexity and slow response speed. This article systematically outlines the application path of artificial intelligence algorithms in power system fault diagnosis and early warning, and proposes a hybrid intelligent diagnosis framework based on multimodal data fusion. Through breakthroughs in three core technologies: physical mechanism embedding, federated learning collaboration, and digital generative deduction, we have achieved a breakthrough in fault recognition accuracy of 99.2%, extended prediction time window to 30 minutes, and reduced positioning error to 3 mm.

    • Research on SOC estimation of power battery: AICDKF-FF algorithm based on thermoelectric coupling model

      2025, 41(20):106-108.

      Abstract (110) HTML (0) PDF 1.59 M (455) Comment (0) Favorites

      Abstract:As one of the core functions of battery management systems (BMS), accurate estimation of the state of charge (SOC) of lithium batteries can effectively avoid battery losses caused by overcharging or overdischarging. Therefore, the accuracy of SOC estimation has always been a research focus. To improve the performance of particle filtering (PF) in SOC estimation, this paper proposes an adaptive iterative center difference Kalman particle filtering (AICDKF-PF) algorithm. This method first optimizes the central differential Kalman filter (CDKF) through adaptive and iterative strategies, and then fuses it with PF to achieve SOC estimation on a joint simulation platform. The simulation results show that under DST and NEDC operating conditions, the average error and root mean square error of AICDKF-PF are reduced compared to other algorithms, and the SOC estimation accuracy is significantly improved.

    • Research on application of AI in non-destructive testing of roads and bridges

      2025, 41(20):109-111.

      Abstract (139) HTML (0) PDF 1.22 M (554) Comment (0) Favorites

      Abstract:As key transportation infrastructure, the safety status of roads and bridges is closely related to peoples lives, property, and socioeconomic development. With the increase of service life and traffic flow, bridge structures are prone to damage such as cracks and rust. Traditional manual visual inspection methods are subjective, inefficient, and have limited detection range, making it difficult to meet the needs of modern bridge inspection. In this context, the application research of AI technology in the field of non-destructive testing of roads and bridges is becoming increasingly important.

    • >System Design
    • Development of automatic inspection system for embankment hazard using “UAV+AI” technology

      2025, 41(20):112-114.

      Abstract (198) HTML (0) PDF 1.20 M (476) Comment (0) Favorites

      Abstract:The article designs and develops a “UAV+AI” automatic inspection system for embankment hazards, aiming to improve the intelligence level and hazard recognition efficiency of embankment safety monitoring. By using a UAV platform equipped with multispectral sensors and high-precision cameras to collect real-time image data of the embankment surface and nearshore waters, an improved deep learning algorithm is adopted to fuse multi-source remote sensing information, and a dynamic monitoring model for the health status of the embankment is constructed. Establish a typical hazard feature database based on historical inspection data, and achieve intelligent recognition of common hazards such as leakage, piping, and cracks by comparing the differences between real-time monitoring data and benchmark features. When abnormal signals are detected, the system quickly determines the type of danger through 3D point cloud reconstruction and semantic segmentation technology, and achieves centimeter level accuracy in danger location based on geographic information systems.

    • Design and implementation of employee online training system based on SpringBoot

      2025, 41(20):115-117.

      Abstract (317) HTML (0) PDF 1.28 M (476) Comment (0) Favorites

      Abstract:Employee skill training is an important means for enterprises to achieve sustainable development. Aiming at the problems of high cost, long time consumption, and low efficiency in traditional training methods, based on Java language, using the “Vue+SpringBoot” architecture and MySQL database, an online training system for enterprise employees is analyzed, designed, and implemented, aiming to improve the efficiency of technical training for enterprise employees. The system mainly includes three functional modules: system management, training management, and user management. Through this system, administrators and teachers can create courses, publish courseware, input test questions, and generate test papers online, while enterprise employees can learn and take exams online.

    • Design of property management system based on B/S architecture

      2025, 41(20):118-120.

      Abstract (261) HTML (0) PDF 1.25 M (491) Comment (0) Favorites

      Abstract:With the acceleration of urbanization and the large-scale development of residential communities, the traditional property management model is facing severe challenges in terms of operational efficiency, information integration, communication and coordination, and service response. To enhance the efficiency of property management, optimize service quality and increase owner satisfaction, this article designs and implements a property management system based on the B/S (Browser/Server) architecture. This system takes a Web browser as the client and realizes the remote access of server-side applications through the Internet or local area network, promoting the transformation of property management towards informatization, networking and intelligence. The background and functional requirements of the project construction were systematically analyzed, the overall architecture design plan of the system was proposed, the core functional modules were clarified, and the design of the key database table structure was completed. This system is designed to provide property management companies with an efficient, convenient and secure management platform, and to build a transparent, interactive and convenient service window for property owners, thereby achieving a win-win situation for both property management companies and property owners.

    • Design of 8-bit RISC-CPU based on FPGA

      2025, 41(20):121-123.

      Abstract (181) HTML (0) PDF 1.26 M (457) Comment (0) Favorites

      Abstract:Based on the Field Programmable Gate Array (FPGA) technology and the hardware description language Verilog HDL, and adopting the top-down design method and modular design concept, the customized design, functional simulation, download verification and system test of the CPU were implemented in the Quartus II integrated environment. The article completed the design of the arithmetic unit (ALU) module, the controller module and the RAM/ROM module using Verilog HDL language. It systematically expounded the CPU design method based on FPGA and conducted a comprehensive verification through the development platform, successfully realizing the basic functional test of the CPU. The results show that based on FPGA technology, flexible customization of CPU cores and large-scale integrated circuits can be achieved. It has good reconfigurability, reliability and scalability, and can adapt to diverse practical application requirements.

    • Research on control system for new energy grid connection without power outage based on wireless communication link

      2025, 41(20):124-126.

      Abstract (171) HTML (0) PDF 1.26 M (535) Comment (0) Favorites

      Abstract:With the transformation of the global energy structure, the access of new energy has become an important part of the power system. This article studies a new energy grid-connected uninterrupted operation control system based on wireless communication links, with a focus on the optimal selection of wireless communication links, the design of uninterrupted operation control strategies and their implementation methods. Meanwhile, the performance of the system in terms of real-time performance, stability and data transmission accuracy was analyzed. Through these designs, the control system can effectively support uninterrupted operation during the grid connection process of new energy.

    • Design and implementation of intelligent monitoring platform for UPS system status in communication room

      2025, 41(20):127-129.

      Abstract (123) HTML (0) PDF 1.19 M (486) Comment (0) Favorites

      Abstract:By analyzing the demand for stable operation of uninterruptible power supply (UPS) systems in communication rooms and combining the shortcomings of existing monitoring technologies, an intelligent monitoring platform solution integrating Internet of Things, big data analysis and other technologies is proposed. The overall architecture, hardware selection, and software functional module design of the platform were elaborated in detail, including data collection, processing, analysis, warning, and visual display. Through actual testing, the platform can accurately monitor the status of the UPS system in real time, effectively improve fault warning and processing efficiency, reduce operation and maintenance costs, and ensure the stable operation of the communication room.

    • Research on portable low voltage leakage monitoring instrument and its monitoring method

      2025, 41(20):130-132.

      Abstract (140) HTML (0) PDF 1.25 M (445) Comment (0) Favorites

      Abstract:Low-voltage leakage monitoris primarily used to detect cable leakage in low-voltage distribution systems, and has a wide range of applications in outgoing cabinets, voltage boxes, and other outgoing equipment. However, traditional monitors are limited by their box structure and cannot effectively store the monitoring wires and power light components used for connection, resulting in problems such as inconvenient carrying, easy damage to accessories, and complex use and operation. In response to the above shortcomings, this article proposes a portable low-voltage leakage monitoring device and its monitoring method. By optimizing the structure of the device body, designing detachable storage components, and rationalizing the operation process, efficie nt storage and convenient portability of the monitoring line and power light components are achieved. This design improves the flexibility, adaptability, and portability of the device in different application scenarios. The actual application results show that the instrument performs excellently in monitoring accuracy and fault point confirmation speed, meeting the practical needs of the power industry.

    • Construction and performance analysis of smart city service recommendation model based on user profile

      2025, 41(20):133-135.

      Abstract (141) HTML (0) PDF 1.25 M (468) Comment (0) Favorites

      Abstract:To meet the needs of convenience and personalized services, this article proposes a user portrait oriented smart city service recommendation model and analyzes its performance. Firstly, based on the data of smart city user groups, a user profile is constructed, and multidimensional labels such as personal attributes and consumption attributes are extracted to form a user profile label matrix. Secondly, by calculating user similarity, predicting ratings, and ranking service items, rating prediction and service recommendation for unknown items of target users can be achieved. The experimental results show that the model can effectively capture users? personalized needs and behavioral characteristics, and improve the accuracy and adaptability of smart city service recommendations.

    • Full process energy consumption prediction and low-carbon scheduling system for steel pipe production

      2025, 41(20):136-138.

      Abstract (124) HTML (0) PDF 1.18 M (459) Comment (0) Favorites

      Abstract:A system solution integrating energy consumption prediction and low-carbon scheduling is proposed for the high energy consumption characteristics of the entire steel pipe production process. By constructing an energy consumption prediction model based on multi-source data fusion and combining it with dynamic optimization scheduling algorithms, real-time energy consumption monitoring and low-carbon path planning of the production process can be achieved.

    • Construction scheme of multimedia information fusion display system in virtual exhibition hall

      2025, 41(20):139-141.

      Abstract (173) HTML (0) PDF 1.19 M (518) Comment (0) Favorites

      Abstract:With the deep integration of digital technology and virtual reality technology, virtual exhibition halls have become the core carrier of cultural dissemination, commercial display, and educational innovation. This article focuses on the construction of multimedia information fusion display system, proposes a system architecture of “technology—content—interaction”, and combines key technologies such as 3D modeling, edge computing, AI algorithm to build a full process solution covering space reproduction, content production, intelligent interaction, and system integration.

    • Research on DCS hardware real-time monitoring and intelligent maintenance system based on PI system

      2025, 41(20):142-144.

      Abstract (140) HTML (0) PDF 1.23 M (481) Comment (0) Favorites

      Abstract:With the continuous deepening of industrial automation, distributed control systems (DCS) have become the core control architecture of modern industrial production processes. However, DCS hardware equipment is prone to problems such as aging, wear and tear, and performance degradation during long-term operation. Once a malfunction occurs, it may lead to production interruption, economic losses, and even safety accidents. As a high-performance process information management platform, the Plant Information (PI) system has powerful real-time data collection, storage and analysis capabilities, providing a new technical path for the monitoring and maintenance of DCS hardware equipment. By building a real-time monitoring platform based on the PI system and integrating it with an intelligent maintenance system, real-time monitoring of the status of key hardware such as controllers, I/O modules, and communication networks, performance degradation prediction, and intelligent maintenance decision-making can be achieved. This significantly reduces unplanned downtime, extends equipment service life, and optimizes maintenance resource allocation. Provide reliable guarantees for the stable operation and economic benefit improvement of industrial enterprises.

    • Design of high-level camera monitoring system for UAV based on computer vision

      2025, 41(20):145-147.

      Abstract (142) HTML (0) PDF 1.20 M (473) Comment (0) Favorites

      Abstract:With the continuous advancement of drone technology, high-level camera monitoring systems equipped with computer vision technology have become effective technical means in modern security monitoring, environmental monitoring, and other fields. This article systematically analyzes the application of computer vision in the high-level camera monitoring system of UAV, focusing on in-depth exploration of key technologies such as real-time image acquisition, object detection algorithm optimization, visual tracking and behavior analysis, data transmission and storage optimization. Based on relevant research and implementation, an efficient and stable architecture design scheme for unmanned aerial vehicle monitoring system is proposed, and the performance advantages and feasibility of the system in practical application scenarios are verified through experiments.

    • Research on intelligent detection instrument and detection method for leakage switch integrity

      2025, 41(20):148-150.

      Abstract (151) HTML (0) PDF 1.39 M (489) Comment (0) Favorites

      Abstract:The leakage switch integrity intelligent detector is a device specifically designed for detecting leakage phenomena, which can effectively monitor the leakage status of electrical equipment. The current leakage detectors on the market have defects such as poor instrument protection and loose interfaces, which affect the accuracy and efficiency of detection. This article innovatively designs an improved intelligent leakage switch integrity detection device, which adopts an automatic opening structure of the box cover and a cable clamping device, improving the convenience, safety, and instrument stability of the detection operation. Experimental verification shows that the detector can achieve efficient and accurate leakage monitoring, effectively prevent the risk of detection interruption caused by loose cables, and has good promotion and application value.

    • Design and optimization of robotic palletizing workstation based on RobotStudio

      2025, 41(20):151-153.

      Abstract (256) HTML (0) PDF 1.28 M (466) Comment (0) Favorites

      Abstract:Aiming at the problems of low design efficiency and long debugging cycle of traditional robot palletizing workstations, this article proposes a design and optimization method for robot palletizing workstations based on RobotStudio. Through the configuration of dynamic conveyor chains, the intelligent grasping strategy of fixtures and the overall logical design of the system, the simulation and verification of the complete workflow of the palletizing robot have been achieved. A systematic analysis and optimization were carried out for the probabilistic failure of grasping that occurred in the simulation. The experimental results show that this scheme can provide a reference for practical engineering applications.

    • Analysis of informatization construction and core capability of industrial cluster infrastructure and data platform under marketization of data element

      2025, 41(20):154-156.

      Abstract (121) HTML (0) PDF 1.25 M (532) Comment (0) Favorites

      Abstract:Based on the investigation of the current situation of digital economy industrial parks, the article points out the main problems existing in the aspects of data supply, collection, storage, management, security and platform construction at present. Following the research path of “problem analysis-platform architecture-core capabilities”, this paper explores the optimization paths for standardizing the service market, enhancing asset liquidity and strengthening data value mining. It summarizes the construction concepts of industrial cluster infrastructure and data development platforms, and conducts a systematic analysis of their framework design and core capabilities, with the aim of providing references for platform construction in relevant regions.

    • Design and implementation of passive message capture network device monitoring system

      2025, 41(20):157-159.

      Abstract (131) HTML (0) PDF 1.34 M (466) Comment (0) Favorites

      Abstract:A passive monitoring system for capturing abnormal information of network devices is designed to address the limitations of traditional network monitoring systems in fault monitoring, as well as the high cost and poor compatibility of commercial systems. The system is based on an open source platform and uses the Simple Network Management Protocol (SNMP) message mechanism to achieve a monitoring mode that combines active polling and passive reception. The system can receive messages sent by network devices and parse them in conjunction with the management information database to promptly obtain device anomaly information. The interface of system is intuitive, providing flexible trigger configuration, can recognize multiple alarm types, and is compatible with multiple network devices. Actual testing shows that the system can generate timely alarm prompts in the face of unexpected events such as device port abnormalities, network topology changes, and protocol session interruptions, significantly improving the monitoring and handling efficiency of network operations and maintenance.

    • Research on load test data analysis and bridge life prediction model for highway bridges

      2025, 41(20):160-162.

      Abstract (123) HTML (0) PDF 1.17 M (506) Comment (0) Favorites

      Abstract:With the rapid development of the transportation industry, highway bridges, as important infrastructure connecting various places, are increasingly concerned about their safety and durability. This article aims to explore the stress conditions of bridges in actual operation through in-depth analysis of load test data of highway bridges, and to construct a bridge life prediction model based on analysis, providing scientific basis for bridge maintenance and management.

    • Design and application of dust collector energy management system based on industrial big data

      2025, 41(20):163-166.

      Abstract (135) HTML (0) PDF 1.38 M (486) Comment (0) Favorites

      Abstract:The article introduces the overall architecture, functional module design, and application research of a dust collector energy management system based on industrial big data. The system adopts a layered architecture, including five parts: perception layer, transport layer, data layer, application layer, and presentation layer. The system function enables real-time monitoring of energy consumption and enables statistical analysis and carbon emission accounting of energy usage. The energy consumption analysis results can be visualized in various ways. By constructing an energy-saving algorithm model and combining it with industrial big data for optimization training, the system can provide energy-saving control strategies for dust removal processes. This system provides strong technical support for environmental protection enterprises to achieve energy conservation, cost reduction, and efficiency improvement. It is a typical case of the integration of industrialization and informatization in the field of environmental protection, and an important direction for future environmental protection enterprises to promote technological innovation and business upgrading.

    • Research on construction of security research and development ecology for mobile applications in power system

      2025, 41(20):167-174.

      Abstract (143) HTML (0) PDF 1.93 M (522) Comment (0) Favorites

      Abstract:The article delves into the issues in the development environment of mobile applications in the power system, including safety isolation, security system, and balance between safety and efficiency. It systematically analyzes the characteristics of mobile applications and the required safety standards. On this basis, a comprehensive laboratory solution is proposed, covering the overall architecture design of R&D protection, security measures, test case design, standardization development, and optimization of laboratory operation and management. The article particularly emphasizes the strengthening of laboratory partition protection strategies, network security design, and terminal security management, and introduces the construction of a mobile application real machine testing platform, striving to achieve a dual improvement of security and efficiency in the research and development process. Finally, specific organizational security measures were proposed from both technical and operational dimensions to improve the technical protection standard system and enhance the security operation and maintenance management mechanism. This study has important practical guidance significance for improving the safety and efficiency of mobile application research and development in the power system.

    • Design and practice of project based teaching intelligent mentor system based on large language model

      2025, 41(20):175-176.

      Abstract (240) HTML (0) PDF 1.16 M (476) Comment (0) Favorites

      Abstract:This article designs a project-based teaching intelligent mentor system based on the big language model, which has functions such as project generation, task allocation, progress tracking, and effectiveness evaluation. The practice in courses such as “data collection and preprocessing” and “data visualization technology” has shown that the system can effectively improve teaching efficiency, meet students? personalized learning needs, and significantly reduce teachers? teaching burden. The research results not only provide new ideas for educational technology innovation and promote high-quality development of education, but also provide strong support for cultivating innovative talents.

    • Construction of quality management system for urban road and traffic engineering based on IoT and cloud computing

      2025, 41(20):177-180.

      Abstract (128) HTML (0) PDF 1.31 M (521) Comment (0) Favorites

      Abstract:In response to the problems existing in urban road and traffic engineering, such as lagging quality management information, untimely feedback, and insufficient coverage of manual inspections, this article constructs an intelligent quality control system based on the IoT and cloud computing. This system utilizes multiple types of sensors to collect real-time data on road structure conditions and the environment. Relying on edge computing and cloud analysis technologies, it has established a closed-loop management mechanism integrating perception, transmission, processing and feedback, achieving intelligent supervision throughout the entire process from construction to operation. The article analyzes the applicability and technological integration potential of IoT and cloud computing technologies in the field of transportation engineering, proposes a four-layer architecture design scheme, and verifies the system deployment effect based on actual project cases. The results show that the system performs significantly in terms of defect response efficiency, risk early warning ability and comprehensive information coverage, and has good practicability and promotion value.

    • Design and implementation of radar clutter simulation software system

      2025, 41(20):181-182.

      Abstract (137) HTML (0) PDF 1.24 M (438) Comment (0) Favorites

      Abstract:Radar clutter is one of the key factors affecting the performance of radar systems, and its simulation is of great significance for system design, testing, and optimization. The article proposes a design and implementation scheme for a computer-based radar clutter simulation software system. This system can efficiently generate various types of radar clutter signals, including meteorological clutter, sea clutter, etc., and supports user-defined clutter parameters. Through modular design and user-friendly interface, this software provides powerful tool support for the research and development of radar systems. The experimental results show that the system can efficiently and accurately generate various types of clutter signals, which can meet the needs of radar system testing and evaluation.

    • Design and implementation of Chinese medicine powder cell microscopic image recognition system based on YOLOv12

      2025, 41(20):183-186.

      Abstract (337) HTML (0) PDF 1.52 M (516) Comment (0) Favorites

      Abstract:Microscopic cell identification of traditional Chinese medicine powder is a key link in the identification of traditional Chinese medicine. Traditional methods rely on manual microscopy, which has the limitations of low efficiency and strong subjectivity. In order to realize automatic recognition, this paper designed and implemented a cell microscopic image recognition system of traditional Chinese medicine powder based on YOLOv12. The system adopts a hierarchical architecture. The model inference layer integrates the latest YOLOv12 object detection algorithm, which is responsible for the accurate positioning and classification of cells. The result display layer builds a Web application based on Vue.js and Flask framework, and uses MySQL database for data management to provide efficient interactive interface for users. Experimental evaluation shows that the average precision (mAP@ 50) of the system on the traditional Chinese medicine powder cell dataset reaches 83.3%, and it has the inference speed to meet the real-time needs of industrialization. The system provides an efficient and accurate automatic solution for the microscopic identification of traditional Chinese medicine, and has positive significance for promoting the intelligent process of quality control of traditional Chinese medicine.

    • >Engineering Practice
    • Technological innovation and application of pipeline sealing design and hot melt adhesive coating process optimization in telecommunication network transformation

      2025, 41(20):187-189.

      Abstract (133) HTML (0) PDF 1.18 M (528) Comment (0) Favorites

      Abstract:In response to the problems of low efficiency and insufficient stability existing in the sealing and hot melt adhesive coating processes of telecommunication network pipelines, this article proposes a series of technological innovation solutions, including methods such as hierarchical heating coating, pipe diameter adaptive sealing structure, and precise positioning and fixation of joints. By introducing a three-stage temperature control coating process, wedge-shaped ring self-compensating sealing rings and three-dimensional coordinate positioning technology, the reliability and construction efficiency of pipeline sealing have been significantly enhanced. Based on the practical application of the telecommunications metropolitan area network renovation project in a certain eastern province, the significant effects of the above-mentioned technologies in reducing maintenance costs and extending the service life of pipelines have been verified, providing a useful technical reference for the highquality upgrade of telecommunications network infrastructure.

    • Application of communication technology in big data network security situation assessment

      2025, 41(20):190-192.

      Abstract (113) HTML (0) PDF 1.23 M (496) Comment (0) Favorites

      Abstract:In response to the complex security threats faced by big data networks, this paper investigates the application of communication technology in network security situation assessment. By analyzing the technical characteristics of mobile communication technology and Narrowband Internet of Things (NB-IoT), combined with typical security issues such as DDoS attacks and data breaches, a communication technology-supported evaluation method is proposed. This method encompasses three key components: calculation of security indicator weights, monitoring of abnormal traffic, and generation of comprehensive scores. Experimental validation demonstrates that the proposed method can effectively improve the accuracy of abnormal traffic detection and enhance device status monitoring capabilities, significantly reducing the risk of data breaches while enabling the quantitative assessment of network security situations.

    • Research on enterprise network security governance system in context of data leakage

      2025, 41(20):193-195.

      Abstract (148) HTML (0) PDF 1.18 M (456) Comment (0) Favorites

      Abstract:In the digital era, data has become a core asset for enterprises. However, frequent data leakage incidents have caused significant losses to enterprises. This article delves into the causes, types, and impacts of data leakage, and, drawing on typical cases, explores the challenges enterprises face in network security governance. Based on the current situation, it proposes strategies for building an enterprise network security governance system, covering aspects such as institutional construction, technical protection, and personnel management. The aim is to provide theoretical support and practical guidance for enterprises to effectively address data leakage risks and ensure network security.

    • Network security risk assessment and protection strategy for industrial control system

      2025, 41(20):196-198.

      Abstract (143) HTML (0) PDF 1.18 M (545) Comment (0) Favorites

      Abstract:With the deep integration of industrial control systems and information technology, the network security threats they face are becoming increasingly severe. This article systematically analyzes the core characteristics of network security risks in industrial control systems, constructs a three-dimensional risk assessment model based on asset identification, threat modeling, and vulnerability assessment, and proposes a three-dimensional protection strategy that includes technical protection, management optimization, and emergency response. Empirical research has shown that this scheme can effectively reduce the risk probability of industrial control systems being attacked by networks and improve the security operation level of critical infrastructure.

    • Research on correlation between computer hardware maintenance and network security

      2025, 41(20):199-201.

      Abstract (133) HTML (0) PDF 1.18 M (406) Comment (0) Favorites

      Abstract:With the acceleration of digitalization, computer hardware, as the physical carrier of network systems, directly influences the stability of cybersecurity through its maintenance quality. This article analyzes three dimensions: the direct threats posed by hardware failures, the indirect impacts of environmental factors on hardware, and the correlation between hardware vulnerabilities and cyber attacks. By integrating the foundational role of hardware maintenance in security technologies such as firewall deployment, intrusion detection, and data encryption, a cybersecurity protection framework based on full lifecycle management of hardware is proposed. The aim is to provide theoretical support and practical guidance for constructing a “software-hardware collaborative” cybersecurity system.

    • Application of IP technology in cloud computing data center networking

      2025, 41(20):202-204.

      Abstract (113) HTML (0) PDF 1.19 M (426) Comment (0) Favorites

      Abstract:As the core infrastructure of modern information technology, cloud computing has strongly driven the rapid development of large-scale data centers. Faced with the demand for efficient processing and fast transmission of massive data, the network architecture of data centers urgently needs to strengthen its ability to achieve high throughput, low latency, and scalability. In this context, efficient and scalable network technology has become the key to data center design and optimization. This article focuses on the various applications of IP technology in cloud computing data center networking, with a focus on system analysis of core issues such as IP infrastructure design, routing and load balancing technology, network virtualization solutions, high availability and fault tolerance mechanisms. The aim is to provide theoretical references and practical guidance for the design and performance optimization of related network architectures.

    • >Cross Fusion
    • Application research of PLC control technology in power plant sample preparation system

      2025, 41(20):205-207.

      Abstract (145) HTML (0) PDF 1.17 M (475) Comment (0) Favorites

      Abstract:The power plant sampling system is a key link in fuel management for coal-fired power plants, and its degree of automation directly affects the representativeness of coal samples and the accuracy of detection results. This article constructs a control system architecture suitable for the sample preparation process based on PLC control technology, achieving precise linkage control of the entire process including sampling, crushing, scaling, mixing, and bagging. Through program architecture design and signal processing analysis, an efficient and stable automatic control logic has been constructed. The running results show that the system runs stably and has good scalability, providing a reference for the automation upgrade of the sample preparation system.

    • Research on application of 5G technology in overloading dynamic checkpoint

      2025, 41(20):208-210.

      Abstract (98) HTML (0) PDF 1.81 M (452) Comment (0) Favorites

      Abstract:A dynamic checkpoint solution based on 5G technology is proposed to address the issues of response delay,“data silos”, and regulatory blind spots that exist in traditional overload control methods. This scheme makes full use of the high bandwidth, low delay and wide connection characteristics of 5G network, and combines intelligent weighing, edge computing and AI image recognition technology to build a real-time and intelligent dynamic monitoring system for overload.

    • Impact of subway train door opening strategy based on simulation analysis on interval evacuation efficiency

      2025, 41(20):211-213.

      Abstract (126) HTML (0) PDF 1.30 M (540) Comment (0) Favorites

      Abstract:Urban rail trains have a large passenger capacity, and in the event of an emergency, it is necessary to evacuate passengers with the highest efficiency. Multiple door opening strategy is proposed for the evacuation of subway train sections, and evacuation simulation model is constructed for analysis. The results indicate that in the case of bidirectional evacuation, opening the first door of the two end carriages can significantly improve evacuation efficiency. In the case of one-way evacuation, opening the first door of the carriage near the communication channel may be the optimal way.

    • Research and application of high-acceleration life test technology for electronic product

      2025, 41(20):214-216.

      Abstract (146) HTML (0) PDF 1.28 M (495) Comment (0) Favorites

      Abstract:The article summarizes the basic concepts, core principles of high-accelerated life testing and its main differences from traditional accelerated testing methods. Based on a specific example of an electronic product, this paper systematically expounds the implementation process of the high-acceleration life test, including the determination methods of the products working limit and failure limit, and conducts mechanism analysis and verification of improvement measures for the faults that occur during the test. Practical experience shows that this test method effectively shortens the product development cycle, significantly enhances the reliability level of the product, and has reference value for the reliability verification work of similar electronic products.

    • Research on dynamic identification method of flexible load response characteristics of pseudo-power plant based on 5G communication

      2025, 41(20):217-219.

      Abstract (150) HTML (0) PDF 1.28 M (479) Comment (0) Favorites

      Abstract:With the large-scale integration of renewable energy into the power system, the role of flexible loads in potential power plants has become increasingly crucial, serving as an important means to cope with power fluctuations and enhance the stability of the power grid. The article proposes a dynamic identification method for the flexible load response characteristics of pseudo-power plants based on 5G communication. By establishing a flexible load response model, designing a dynamic identification algorithm, and relying on 5G communication technology to optimize the data acquisition and transmission process, the real-time identification and system optimization of the flexible load response characteristics have been achieved. Finally, the effectiveness of this method was verified through actual cases.

    • Research on voltage control technology for distributed photovoltaic substation based on centralized configuration of reactive power flow and energy storage

      2025, 41(20):220-222.

      Abstract (158) HTML (0) PDF 1.19 M (432) Comment (0) Favorites

      Abstract:With the increasing penetration rate of distributed photovoltaics in the distribution network, the intermittency and fluctuation of its output have a significant impact on the voltage stability of the substation area. This article proposes a distributed photovoltaic substation voltage control strategy based on centralized configuration of reactive power flow and energy storage. By constructing a collaborative control model between the energy storage system and the photovoltaic inverter, combined with reactive power flow optimization algorithm, dynamic regulation of substation voltage is achieved. Simulation and experimental results show that this strategy can effectively suppress voltage fluctuations, improve the voltage qualification rate to over 98%, and reduce line losses by about 15%, providing theoretical support and technical solutions for voltage control in high penetration distributed photovoltaic access.

    • Optimization of DC bus architecture and energy scheduling strategy for integrated photovoltaic storage and charging power station

      2025, 41(20):223-225.

      Abstract (254) HTML (0) PDF 1.19 M (501) Comment (0) Favorites

      Abstract:With the rapid development of new energy and energy storage technologies, integrated photovoltaic , energy storage, and charging stations have gained significant attention from both industry and academia as a crucial platform for the efficient utilization of clean energy. This article focuses on the optimal design of the DC bus architecture within such integrated stations and proposes an optimization method based on system topology and power flow characteristics. Furthermore, a dynamic energy scheduling strategy is designed to coordinate the operation among PV generation, the energy storage system, and charging loads. Simulation results demonstrate that the optimized architecture significantly reduces system power loss, enhances operational stability, and improves energy utilization efficiency through intelligent scheduling, thereby meeting the demands of multiple loads under various complex operating conditions.

    • Research on comprehensive carrying capacity of distribution network adapted to large-scale access of distributed photovoltaics

      2025, 41(20):226-228.

      Abstract (143) HTML (0) PDF 1.19 M (510) Comment (0) Favorites

      Abstract:With the advancement of the “dual carbon” target, the scale of distributed photovoltaic installed capacity has shown explosive growth. The accessible capacity of the distribution network in central and eastern provinces of China is approaching saturation, and the problem of insufficient carrying capacity is becoming increasingly prominent. This article constructs a comprehensive carrying capacity evaluation system for distribution networks from the perspectives of technology, economy, and system friendliness, and proposes an improvement path based on the collaborative optimization of “source network load storage”. Through IEEE 33 node system simulation verification, the adoption of dynamic carrying capacity grading warning mechanism and energy storage optimization configuration scheme has improved the acceptance capacity of distributed photovoltaics in the distribution network by 23.6%, and the voltage qualification rate has increased to 99.2%, verifying the effectiveness of the model. The research provides theoretical support for solving the contradiction between the largescale access of distributed photovoltaics and the lagging transformation of distribution networks.

    • Exploration of archive data management ideas in context of big data governance environment

      2025, 41(20):229-231.

      Abstract (113) HTML (0) PDF 1.18 M (471) Comment (0) Favorites

      Abstract:This article combines the actual situation of archive data management in Qingyuan Rehabilitation Center for Persons with Disabilits, analyzes the impact of big data governance on archive data management, explores the current problems in archive data management, and proposes corresponding management ideas, aiming to improve the quality and efficiency of archive data management in rehabilitation centers and provide better data support for disabled rehabilitation.

    • Research ondual dimensional improvement path of efficiency and compliance of SPD mode in intelligent transformation of scientific research management

      2025, 41(20):232-234.

      Abstract (123) HTML (0) PDF 1.19 M (454) Comment (0) Favorites

      Abstract:Based on the SPD(Scientific Project Digitalization) model, this article explores the path of intelligent transformation in scientific research management, and proposes a method to achieve dual improvement of efficiency and compliance in scientific research project initiation, fund management, achievement evaluation, and compliance review through digital and intelligent tools. Through case analysis and practical application, the results show that the SPD model can significantly reduce the complexity of scientific research management processes, improve approval efficiency, and strengthen compliance control, providing replicable management optimization solutions for scientific research institutions.

    • Optimization and scheduling strategy for distributed energy in smart grid

      2025, 41(20):235-237.

      Abstract (185) HTML (0) PDF 1.18 M (430) Comment (0) Favorites

      Abstract:With the continuous growth of global energy demand and the accelerated transformation of energy structure, the penetration rate of distributed energy in smart grid has significantly increased. With the advantages of decentralization, renewability and operation flexibility, distributed energy provides important support for the efficient and stable operation of smart grid. However, its output volatility and uncertainty increase the complexity of power grid dispatch. This paper focuses on the problem of optimal scheduling of distributed energy in smart grid. Based on the systematic analysis of the characteristics of distributed energy, the scheduling requirements of smart grid and the limitations of existing methods, a series of improved optimal scheduling strategies are proposed to improve the reliability, economy and renewable energy absorption capacity of grid operation.

    • Rapid inventory method based on UHF RFID reader

      2025, 41(20):238-240.

      Abstract (112) HTML (0) PDF 1.26 M (420) Comment (0) Favorites

      Abstract:In recent years, ultra-high frequency RFID technology has developed rapidly and has been widely applied in many fields such as asset management and supply chain management. However, according to international standards, UHF RFID readers generally use ALOHA-type algorithms to handle multi-tag identification. When the number of tags is large, the inventory efficiency drops significantly. To address this issue, the article proposes an improved dynamic frame slot ALOHA algorithm at the system design level to achieve rapid and efficient label inventory. The protocol design principle and software implementation process of this algorithm are elaborated emphatically. Through actual tests, it is verified that it can significantly improve the inventory efficiency in multi-label scenarios, and has good practicability and simplicity while meeting the performance requirements. This research achievement has certain reference value for the development of Internet of Things devices adopting ALOHA-type protocols.

    • High precision remote sensing image analysis based on point cloud data processing system

      2025, 41(20):241-243.

      Abstract (97) HTML (0) PDF 1.87 M (563) Comment (0) Favorites

      Abstract:Remote sensing image analysis technology is widely used in fields such as geographic information systems, environmental monitoring, and urban planning. The article proposes a remote sensing image analysis method that combines point cloud data processing system, aiming to improve the accuracy of image classification, object detection, and change detection. The experimental results show that this method can significantly improve the classification accuracy of remote sensing images and demonstrate good adaptability and reliability in multiple typical application scenarios.

    • Implementation of modular exponential algorithm based on Fork/Join mode

      2025, 41(20):244-246.

      Abstract (154) HTML (0) PDF 1.45 M (513) Comment (0) Favorites

      Abstract:Public key encryption and digital signature algorithms generally rely on modular exponentiation, and the core bottleneck of modular exponentiation of large integer is multiplication and exponentiation with high computational complexity. Based on the existing research, this paper systematically analyzes the application characteristics of binary method, Karatsuba algorithm and Fast Fourier Transform (FFT) in modular exponentiation, and optimizes them for computational efficiency and resource consumption. Furthermore, based on Java language environment, the improved algorithm is implemented by using multi-thread Fork/Join framework to improve parallel computing performance and overall computing efficiency.

    • Application and value of prototype design in software development: from concept to implementation

      2025, 41(20):247-249.

      Abstract (124) HTML (0) PDF 1.33 M (512) Comment (0) Favorites

      Abstract:With the rapid development of information technology, the software development process is also continuously evolving. As a bridge connecting creativity to actual products, prototype design plays a crucial role. This article delves into the concept, purpose, implementation methods of prototype design, and its significance for different development positions, showcasing the application of prototype design in real projects through specific case studies.

    • Mechanical calculation and analysis of hydraulic support pin shaft based on Internet of Things technology

      2025, 41(20):250-252.

      Abstract (154) HTML (0) PDF 1.37 M (475) Comment (0) Favorites

      Abstract:As the core load-bearing component of the support structure, the mechanical properties of the hydraulic support pin directly affect the overall safety and stability of the equipment. With the increasing demand for highreliability support structures in large-scale industrial fields such as nuclear power, thermal power, and steel, the precise force analysis of pins has become particularly important. Based on Internet of Things technology, this article conducts an in-depth study on the force state of the pin shaft of the hydraulic support through real-time monitoring by sensors and numerical calculation methods. By constructing the mechanical model of the pin shaft and deriving its force calculation formula, the model was effectively verified in combination with experimental data. Compared with traditional static analysis methods, this article utilizes the Internet of Things platform to collect pin shaft stress data in real time, significantly improving the accuracy of calculation and response timeliness. The experimental results show that this method can effectively improve the monitoring accuracy, optimize the design of the support structure, and help extend the service life of the pin shaft.

    • Accurate prediction method for offshore platform power load based on data analysis

      2025, 41(20):253-255.

      Abstract (154) HTML (0) PDF 1.20 M (425) Comment (0) Favorites

      Abstract:To solve the problems of data complexity, heterogeneity, and load fluctuation in offshore platform power load forecasting, this article proposes a power load forecasting method based on the classification and regression tree (CART) algorithm to improve the accuracy and real-time performance of forecasting. Firstly, analyze the characteristics, influencing factors, and prediction difficulties of the power load on offshore platforms. Secondly, construct a load forecasting data analysis model based on CART and design a complete forecasting process. Finally, the model was validated through experiments. The results show that this method can effectively capture the load variation pattern, with a small prediction error and an average relative error of 5.86%. It has high feasibility and practical value.

    • Adaptive sliding mode control of interval type-2 fuzzy system based on observer

      2025, 41(20):256-259.

      Abstract (164) HTML (0) PDF 1.35 M (489) Comment (0) Favorites

      Abstract:The article investigates the adaptive sliding mode control problem of interval type-2 fuzzy systems with external disturbances. A new sliding mode control method based on observer is designed for fuzzy systems with different input matrices, which can ensure the consistent ultimate boundedness of the closed-loop system. The feasibility of the design method is verified through numerical simulation.

    • Analysis of reason for refusal of payment in intelligent review of medical insurance in third tier hospital and improvement countermeasure

      2025, 41(20):260-262.

      Abstract (213) HTML (0) PDF 1.16 M (502) Comment (0) Favorites

      Abstract:Taking the case of medical insurance refusal under the intelligent review system of a tertiary hospital as the research object, this study analyzes the reasons for medical insurance fee refusal, aiming to improve the compliance of medical behavior and ensure the security of medical insurance funds through data-driven methods. Based on the intelligent audit deduction data of the entire hospitals medical insurance from January to December 2024, combined with medical insurance policies and clinical cases, descriptive statistical analysis was conducted using Excel. The results indicate that hospitalization is a high-risk area for medical insurance violations, with main issues including medication beyond the scope of medical insurance payment, suspected over examination, duplicate charges, over standard charges, and over indication medication. These situations account for a relatively high proportion of intelligent audit deductions. The research results have practical significance for identifying risk links in the supervision of medical insurance funds, optimizing internal cost control and quality management in hospitals, and providing reference for medical institutions to adapt to the trend of intelligent supervision of medical insurance.

Current Issue


Volume , No.

Table of Contents

Archive

Volume

Issue

Most Read

Most Cited

Most Downloaded