首页|结合自然语言处理与知识图谱的电力项目安全管理应用设计

结合自然语言处理与知识图谱的电力项目安全管理应用设计

扫码查看
电力项目安全管理需要处理大量的文本数据,传统方法在处理效率和准确性上存在一定不足.因此,研究结合自然语言处理构建电力项目安全管理知识图谱.知识图谱验证结果表明,研究的知识抽取模型表现出色,准确率、召回率和F1值分别为0.94、0.91和0.92.知识图谱的知识更新频率为849条/天,知识库扩展速度为857条/天,检索速度平均为1 048次/秒,能有效降低安全隐患.实际应用结果表明,引入知识图谱后,安全问题解决时间最低降至18.7分钟,决策准确性平均提升至91.46%,人力资源和设备利用率显著提高,分别为91.37%和88.64%.用户满意度调查显示总体满意度高达9.58.研究为电力项目安全管理者提供更为先进、全面的管理工具,对电力产业的可持续发展具有重要意义.
Design of Power Project Safety Management Application Integrating Natural Language Processing and Knowledge Graph
Safety management in power projects involves dealing with a substantial amount of textual data,and traditional meth-ods exhibit limitations in terms of efficiency and accuracy.Hence,this research integrates natural language processing to construct a knowledge graph for power project safety management.The results demonstrate outstanding performance of the proposed knowledge ex-traction model,with accuracy,recall,and F1 score reaching 0.94,0.91,and 0.92,respectively.Upon introducing the knowledge graph,the minimum time for resolving safety issues is reduced to 18.7 minutes.The average accuracy of decision-making improves to 91.46%,and there is a significant enhancement in the utilization of human resources and equipment,reaching 91.37%and 88.64%,respectively.The knowledge graph exhibits a knowledge update rate of 849 items per day,a knowledge base expansion speed of 857 items per day,and an average retrieval speed of 1048 times per second.A user satisfaction survey indicates an overall satisfaction rate of 9.58.This research provides power project safety managers with more advanced and comprehensive management tools,holding significant implications for the sustainable development of the power industry.

power projectsafety managementnatural language processingknowledge graph

戴玉艳、章瑶易、安佰龙、陆柳

展开 >

国网上海市电力公司电力营销服务中心(计量中心),上海 200030

电力项目 安全管理 自然语言处理 知识图谱

国网上海科技项目

52090D230004

2024

自动化与仪器仪表
重庆工业自动化仪表研究所,重庆市自动化与仪器仪表学会

自动化与仪器仪表

CSTPCD
影响因子:0.327
ISSN:1001-9227
年,卷(期):2024.(8)