In order to improve the accuracy and efficiency of power system network security risk assessment,this paper proposes a method of power system network security risk assessment based on RBF neural network.Firstly,establish a security risk assessment index system,and combine it with the risk assessment matrix.By calculating the information entropy of the index,the risk level of the evaluation index is determined and divided into five risk levels.Finally,the network security risk distribution fusion eigenvalues are calculated and input,and based on the network security risk assessment results of power system,RBF neural network is used to construct the network security risk assessment model of power system.Experiments have shown that the method achieves the desired goal,96％accuracy in the assessment of the highest security risk,and 98％efficiency in the assessment.
2023年11月2日至3日,"CPEM第4届电力人工智能大会暨第2届电力行业数字化转型大会、PD2023全国配电数字化技改大会"在贵州省贵阳市召开. 大会由CPEM全国电力设备管理网、国家能源智能电网(上海)研发中心、复杂能源系统智能计算教育部工程研究中心、中国电子劳动学会双碳和能源创新工作委员会、中国人工智能学会智慧能源专业委员会、《电力大数据》期刊、人工智能生态产业协作平台、青岛鼎信通讯股份有限公司联合主办,国电南瑞科技股份有限公司、南京国电南自电网自动化有限公司支持,《中国电力》杂志社提供学术支持,数智云(无锡)工业科技有限公司、上海共燊新科技有限公司、无锡共燊会展服务有限公司承办.
The digital transformation of the power grid aims to achieve the sharing,collaboration,and transmission of data,knowledge,and information across all aspects,breaking down boundaries between departments,humans and machines,as well as management and control.This ultimately maximizes the efficiency of standardized processes under comprehensive global coordination.This paper primarily discusses three fundamental objectives in the construction of a digitalization platform for power grid standards:unifying and standardizing data sources,reshaping and optimizing standard business processes,and promoting visualized embedded applications and efficiency assessment of standards.Through analysis in four aspects:business architecture design,application architecture design,data architecture design,and technology architecture design,a digitalization platform for power grid standards is constructed,enabling collaborative management of standards throughout their lifecycle,efficient retrieval of standard content information,interactive online editing and reading of standards,and laying the foundation for the efficient operation of asset lifecycle management and end-to-end business processes in the future.
The optimization and configuration of power capacity in microgrids are crucial initial steps in microgrid planning and construction.This process requires rational allocation of power sources within the microgrid,considering meteorological data,geographic location,and load demands while meeting economic and reliability constraints.This paper focuses on an independent microgrid in a plateau region and proposes an optimization method for configuring power capacity based on typical meteorological years and time-shiftable oxygen-dependent loads.The algorithm takes into account various constraints,including the ratio of installed renewable energy capacity,power supply reliability,and oxygen supply reliability.The objective is to minimize the annual average cost over the microgrid's entire lifecycle,and a genetic algorithm is used to solve this optimization model.The paper analyzes the results of different configurations and validates the feasibility of the proposed method.This research provides guidance for the planning and construction of independent microgrids,contributing to efficient energy utilization,cost-effectiveness,and improved reliability of power and oxygen supply.
With the rapid growth of industrial and residential electricity consumption in rural and mountainous areas,under-voltage issues are becoming significant and particularly prominent in some periods due to seasonal load and population migration.It is extremely difficult to achieve a balance between input and benefit when considering adopting traditional transformation measures to solve the above problems.Therefore,two types of grid transformation measures,AC-DC series/parallel connection,are proposed to solve under-voltage issues owing to excessive power supply distance.Firstly,this paper introduces basic situations of two low-voltage distribution networks(LVDNs)before the two-type transformations,and analyzes the problems of the traditional transformation measures.Secondly,the two-type transformations of AC-DC series/parallel connection are introduced in detail from three aspects:construction scheme,converter equipment and equipment installation.Moreover,the effectiveness of the AC-DC series/parallel connection schemes is verified by analyzing the laboratory test data and applications,and the advantages and disadvantages of the two types of transformations are analyzed.Finally,the effects of the the two types of transformations and the future development direction are summarized.
Harnessing the full potential of flexibility resources in the power system can effectively enhance the integration capacity of renewable energy sources.Carbon capture power plants not only reduce carbon emissions in the power system but also offer potential flexibility for grid regulation.This paper establishes a flexible regulation model for demand response and carbon capture power plants,presents a multi-timescale scheduling approach that considers the flexible operation of carbon capture power plants and demand response,and applies it to the dispatch operation of the power system to minimize the total operational cost.In the day-ahead scheduling phase,demand response and flexible regulation of carbon capture power plants are utilized to shift load demands,achieving the goal of low-carbon economic operation.In the intraday scheduling phase,a predictive control model is employed to adjust the day-ahead scheduling plan to ensure real-time power balance.The results demonstrate that demand response and flexible operation of carbon capture power plants can increase the integration of renewable energy sources in the system while reducing total costs by 18.7％and 1.4％,respectively.Coordinated operation of carbon capture power plants and demand response can reduce the system's total cost by 20.1％,yielding significant benefits.
In order to meet the near real-time monitoring requirements for wildfires in high-risk areas along power transmission lines,which demand low false negatives,high accuracy,wide coverage,and high timeliness,this paper proposes a wildfire detection algorithm based on multi-channel convolutional neural network(MC-CNN)using geosynchronous orbit satellite imagery as the foundation.To some extent,the algorithm reduces the false negative rate of wildfire detection by combining the OTSU algorithm and context algorithm to enhance potential fire points.The principal component analysis(PCA)algorithm is introduced to optimize input features,constructing a multi-channel network structure.Weighted average of different channel fire point recognition weights is determined using a combination of probability and particle swarm optimization(PSO)parameter optimization algorithm,ultimately identifying the fire points.Furthermore,fixed high-temperature heat sources and solar flares are utilized to eliminate false fire points,reducing the false alarm rate.To validate the effectiveness of the proposed algorithm,historical satellite-monitored wildfire cases near power transmission lines between 2019 and 2022 are randomly selected.Known fire point samples are used to validate the results of fire point inversion.The calculation results show that the accuracy of this algorithm for wildfire detection reaches 89.4％.
Seasonal population movements lead to short-term increases in electricity demand,causing overload operation and even failure,such as burnout,in 10kV oil-immersed distribution transformers,resulting in customer complaints.This paper,based on the overload situations of distribution transformers in the Guizhou region,comprehensively considers factors affecting overload capacity,risks associated with overload operation,and insulation life losses.It summarizes the fundamental characteristics of overload operation for distribution transformers and proposes an approach and method to verify their overload capacity through overcurrent and temperature rise testing.The feasibility of this method is verified through on-site testing,providing a basis for addressing overload operation issues promptly.Considering the advantages of vegetable-based insulating oil,such as good electrical properties,high flash point,wide availability of raw materials,and renewability,the paper also validates the replacement of mineral insulating oil with vegetable-based insulating oil to enhance overload capacity.Load performance before and after the oil replacement is analyzed,and recommendations are provided for short-term,medium-term,and long-term measures.
Power plants have a large number of equipment,diverse types,and a significant amount of fault information with complex data coupling relationships.Leveraging knowledge graphs to integrate fault information and develop intelligent applications is particularly helpful for power plants in analyzing equipment operation and fault data.Extracting relationships between entities from massive and heterogeneous data is a key step in building knowledge graphs.This paper presents a design scheme for a relationship extraction tool aimed at constructing a knowledge graph for key power plant generation equipment faults.The tool visualizes fault information and the information processing process,while also allowing user participation in the relationship extraction process.Through iterative training,the relationship extraction model is continuously optimized to improve accuracy.Finally,using real power plant equipment fault data for validation testing,the tool demonstrates a significant improvement in relationship extraction accuracy.Therefore,by constructing high-quality knowledge graphs,we can better manage and maintain power plant equipment,providing essential support for the stable operation of power plants.
Infrared testing(IRT)has experienced rapid growth in recent years,and the technology surrounding IRT has become an important research direction in various fields.This paper provides a comprehensive analysis of the literature on IRT technology from both domestic and international sources,summarizing the future research hotspots in this field.Using the Web of Science(WOS)core collection database,the CiteSpace data mining and analysis platform is utilized to conduct an in-depth analysis of the latest developments in this research area.Multiple dimensions,including publication volume,countries,authors,research institutions,keywords,cited references,and journals,are explored to comprehensively understand the research hotspots and development trends in this field.This paper specifically tracks research directions related to defect detection in metal components,defect detection in composite materials,and infrared image insulation detection technology.
In this paper,we propose a distributed stress detection method for overhead power transmission lines based on the LSTM algorithm.The method utilizes communication fibers in the fiber composite overhead phase line and records fiber gratings to enable transmission capabilities,transforming the distributed stress monitoring problem into a model regression problem.Considering the linear distribution characteristics of fiber sensors,LSTM is used for feature learning and training to obtain the distributed stress sensing monitoring model.Experimental results demonstrate that this model can effectively detect the distributed stress of the conductor,with a root mean square error of 1.023.The average detection error for stress magnitude is 2.032N,and the average detection error for stress position is 1.41mm.Compared to results obtained using ELM and SVM methods,this approach shows improvement and provides a new direction for distributed stress monitoring of fiber composite overhead phase lines.
Joint inspection technology is an important application of artificial intelligence in substations and plays a key role in promoting the intelligent development of substations.State Grid Wenzhou Power Supply Company has introduced advanced technologies such as artificial intelligence-based joint inspection for the intelligent transformation of a certain substation.This paper firstly introduces the overall architecture and key technologies of the AI-based joint inspection system in the substation.Secondly,the functional applications of the system are summarized,including video monitoring technology,robot inspection technology,intelligent linkage technology,infrared temperature measurement technology,image recognition technology,3D reality modeling,one-click sequential control technology,and expert systems.Furthermore,the intelligent operation and maintenance of the substation are preliminarily discussed.Lastly,the operational analysis of the application case in the substation is presented.The results show that the system has achieved the initial substitution of manual labor with machines,effectively improving the safety and production efficiency of the substation.
六盘水供电局始建于1966年11月,隶属于贵州电网有限责任公司,主要负责贵州西部电网——六盘水区域电网的运行管理和整个六盘水地区的电力供应.近年来,六盘水供电局先后荣获"全国文明单位",南方电网公司"文明单位""安全文化示范单位""先进基层党组织""安全生产先进集体""信息化工作先进单位""普法工作先进单位""党建思想政治工作研究先进单位",贵州省"科学技术重大贡献奖",贵州省首届"网络安全技能竞赛"第二名,贵州省国资委"先进基层党组织",贵州电网公司"优秀企业""党风廉政建设先进单位""优秀基层党组织"等多项荣誉.
The tower system,including the tower and bolts,is a critical component for the normal operation of wind turbines.Therefore,accurately identifying potential safety hazards like cracks is crucial.However,the challenges of low distinguishability,poor contrast,and subtle characteristics of cracks have posed difficulties in practical applications.To solve this problem,this paper proposes an improved YOLOv7-SEAttention algorithm model,an advancement on the YOLO series.The enhanced model is compared with various algorithms such as Faster R-CNN,RFCN,SSD,YOLOv5 and YOLOv7.The improved algorithm's recall,precision,and average precision(AP)are comprehensively evaluated.The YOLOv7-SEAttention model exhibits significant advantages in detecting surface cracks in tower systems.Compared to the original YOLOv7 and other models,it achieves a 2.6％increase in AP for tower crack detection,reaching 83.7％,and a 4％increase in AP for bolt crack detection,reaching 84.3％.The results indicate that the improved model presented in this paper can detect surface cracks in tower systems with greater accuracy.
Knowledge graphs hold significant value in the field of knowledge visualization.However,the real-time nature of graph data and the continuous increase in data volume often lead to redundant data,directly impacting the quality and usability of knowledge graphs.To solve this problem,this paper designs a dynamic update method for knowledge graphs applied in the field of power generation.Utilizing the robust graph processing capabilities of the Neo4j graph database,the method achieves efficient updates of knowledge graph data,significantly reducing the occurrence of duplicate nodes and thereby greatly enhancing update accuracy.Dynamic updates of the knowledge graph in the field of power generation equipment ensure that the information remains up-to-date,providing strong support for practical applications.Particularly in the maintenance and optimization of power generation equipment,this timely updated knowledge graph becomes a highly valuable tool.
The goal of full awareness of the power grid's state implies the widespread application of internet of things(IoT)technology in the power grid,where a large number of IoT terminals will be connected to the network through sensing,communication,and computer technologies.Power IoT terminals are characterized by a large quantity,wide geographical distribution,and complex data collection,making them susceptible to breaches and intrusions by attackers.Additionally,traditional centralized authentication systems suffer from issues like single-point failure and performance bottlenecks.This paper,based on blockchain technology,develops and applies technologies such as decentralized identifiers(DID)digital identity based on terminal identifiers,a blockchain-adaptive DID resolver,zero-knowledge proofs based on terminal credential information,and device management based on machine learning algorithms.These technologies culminate in the design of a blockchain-based security authentication system for power Internet of Things terminals.The system has been successfully applied in the power grid,achieving decentralized authentication of IoT devices,reducing network security risks associated with device access,lowering the costs of centralized infrastructure construction and maintenance,and enhancing the efficiency of operations and maintenance personnel.
Equipment quality is fundamental to ensuring the safety and stability of the power grid.Improving the level of grid equipment,advancing towards mid-to high-end equipment,and persistently enhancing the intrinsic safety level of the power grid are top priorities in current and future work.Only by ensuring the quality of grid equipment at the entry point,namely starting from procurement quality,can equipment quality be fundamentally improved.In recent years,State Grid Corporation has conducted verification of supplier qualifications and capabilities before tendering.By verifying and confirming information such as the qualifications,performance,and on-site actual production situation of suppliers,State Grid has gained preliminary insights into whether potential suppliers of power equipment have the qualifications and capabilities to produce qualified products.However,there are still cases of major equipment failures occurring during operation.Therefore,in order to effectively prevent and control technical risks of products,promote the transformation of power enterprises from scale expansion to quality and efficiency-oriented development,strengthen the whole-process quality control of the power grid,and improve equipment quality from the source,this paper first analyzes advanced experiences of domestic and foreign power enterprises,with a focus on understanding their supplier selection and management methods in the procurement process.Based on the actual situation of State Grid,a supplier selection mechanism based on PSO-ELM is established to drive the power grid equipment towards the mid-to high-end level and enhance the intrinsic safety and reliability of the power grid.
Currently,the delivery of secondary curcuit designs in substations still relies on traditional drawing methods.This approach requires personnel with specialized knowledge for drawing verification and field operation and maintenance,which is not conducive to the development of digital technologies in power systems.To achieve intelligent verification of secondary circuit wiring in substations,this paper draws inspiration from the IEC61850 logical modeling approach and proposes a digital twin-based modeling scheme.Starting from the perspective of substation engineering design,the paper performs digital modeling of general layout,screen position arrangement,panel grouping scheme,and crucial wiring designs.Building on this,typical design specifications are transformed into a digital standard rule library,serving as the basis for intelligent verification.By comparing and matching object instance attribute data obtained from scanning on-site cabinet and cable QR codes,this method verifies the design product and aids in construction and later operation and maintenance.Simulation tests on a certain expansion project demonstrate that this method effectively improves the work efficiency of debugging and operation and maintenance personnel.
With the advancement of technology and the progress of society,data exchange in power information networks will become more open,and the application scenarios of power grid control systems will become more complex.Without the guarantee of information network security,the construction of ubiquitous power Internet of Things will be impossible.In light of the current situation of power grid security,this study aims to explore a security strategy for communication information networks in the digital power grid based on cloud-edge collaboration.Firstly,through extensive literature analysis,surveys,and other methods,the main security threats and protection requirements faced by communication information networks in the digital power grid are identified.Secondly,based on this foundation,a security strategy centered around cloud-edge collaboration is proposed,which includes multi-layered solutions such as resource sharing,real-time monitoring,and data encryption.Then,this paper evaluates the strategy from three perspectives:security,reliability,and usability,and verifies its effectiveness and feasibility through simulation experiments.The conclusion indicates that the security strategy for communication information networks in the digital power grid based on cloud-edge collaboration can effectively enhance the security and stability of the digital power grid,and it possesses strong practical application value.
To prevent early weak faults in rolling bearings from evolving into severe faults,this paper proposes a new fault detection method.Initially,principal component analysis(PCA)is employed for feature selection of vibration signals to reduce data dimensions,simplify the structure of vibration data,and enhance feature expressiveness.Then,the complete ensemble empirical mode decomposition with adaptive noise algorithm(CEEMDAN)is used to decompose weak fault vibration signals interfered by background noise.By introducing adaptive noise on top of empirical mode decomposition(EMD),this method enhances the recognition of weak features,separating trend and noise data to improve the accuracy of fault diagnosis.Finally,the transformer model is incorporated to further optimize feature extraction and representation,achieving efficient processing of long sequence data for weak fault feature extraction and characterization.This comprehensive approach,with its advantages in dimension reduction,noise suppression,and long sequence processing,holds promise for significant achievements in fault detection of rolling bearings.
