基于AI和RPA技术的电力生产异常数据智能识别方法  

IntelligentRecognition Method For Abnormal Data in Power Production Based on AI and RPA Technology

作　　者：林轩安 肖宁 王雯婷 邹文强 郭天昀 LIN Xuan’an;XIAO Ning;WANG Wenting;ZOU Wenqiang;GUO Tianyun(Guiyang Power Supply Bureau,Guizhou Power Grid Co.,Ltd.,Guiyang,Guizhou 550002,China)

机构地区：[1]贵州电网有限责任公司贵阳供电局,贵州贵阳550001

出　　处：《长江信息通信》2025年第2期86-88,共3页Changjiang Information & Communications

摘　　要：在电力生产过程中,由于数据来源多样且实时性要求高,导致信息重叠度较高,因此设计一种基于AI和RPA技术的电力生产异常数据智能识别方法。利用RPA自动化采集电力生产数据,通过高效的数据抓取与流程优化,确保电力服务调度和抢修调度数据的精准监控。基于AI技术的深度学习网络,建立电力生产异常数据融合业务模型,通过智能分析服务调度与抢修调度的数据,实现无缝衔接。该模型利用机器学习算法对复杂异常特征进行分类,并通过计算长度特征来区分不同类型的异常数据,形成差异特征集合,最终确定最优的异常分类结构,实现了对异常数据的智能识别与高效管理。实验结果表明,设计方法信息重叠度在了8%以内,低于文献[1]的29%和文献[2]的24%,证明了该方法在电力生产异常数据识别领域中的高效性与精准性。In the process of power production,due to the diverse data sources and high real-time requirements,there is a high degree of information overlap.Therefore,a smart identification method for abnormal data in power production based on AI and RPA technology is designed.Utilizing RPA to automate the collection of power production data,ensuring accurate monitoring of power service scheduling and emergency repair scheduling data through efficient data capture and process optimization.Based on AI technology,a deep learning network is established to establish a business model for the fusion of abnormal data in power production.Through intelligent analysis of service scheduling and emergency repair scheduling data,seamless connection is achieved.This model utilizes machine learning algorithms to classify complex abnormal features,and distinguishes different types of abnormal data by calculating length features,forming a set of differential features,and ultimately determining the optimal abnormal classification structure,achieving intelligent recognition and efficient management of abnormal data.The experimental results show that the information overlap of the design method is within 8%,which is lower than the 29%in reference[1]and the 24%in reference[2],demonstrating the efficiency and accuracy of this method in identifying abnormal data in power production.

关 键 词：AI技术 RPA技术 电力生产 异常数据 智能识别 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]