摘要
针对现有技术台区信息采集效果不佳、台区特征识别能力差等问题,提出了新型的配电台区信息分析方法。该方法采用配电台区全业务标签体系构架实现多源数据融合和构建,通过配网台区画像架构设计,构建出新型的资产设备管理的周期画像。以营配业务为核心建立台区标签体系,实现各种问题分析的模型架构,并借助于标签化识别方式实现台区特征信息的识别。该研究还通过融合多维类别特征识别的方式进行业务场景画像分析,实现复杂维度业务场景画像构建。试验表明,该研究的方法误差精度低,提高了配电台区全业务管理能力。
On the basis of the problems of the poor performance of the existing technology in collecting information in the courts and the poor ability of identifying the court’s characteristics,a new method for analyzing information in the distribution areais proposed.This method adopts the whole service label system architecture of the distribution area to realize the integration and construction of multi-source data.A new cycle portrait of asset equipment management is constructed through the design of the distribution network area portrait architecture.An area labeling system with the distribution business is established as the core to realize the model structure for various problem analysis,and the identification of courts characteristic information with the help of labeling identification.This research also analyzes the business scene profile by integrating multi-dimensional category feature recognition to realize the construction of complex dimensional business scene portrait.Experiments show that the method of this study has low error accuracy and improves the overall business management capability of the distribution area.
作者
秦丹丹
郑高峰
刘丽
韩学民
梁晓伟
孙伟
QIN Dan-dan;ZHENG Gao-feng;LIU Li;HAN Xue-min;LIANG Xiao-wei;SUN Wei(Anhui Electric Power Co.,Ltd.,Hefei 230000,China;Anhui Electric Power Research Institute,Hefei 230000,China)
出处
《信息技术》
2021年第6期136-142,共7页
Information Technology
关键词
特征识别
配电台区
全业务标签体系
业务场景画像
多维类别特征
feature recognition
distribution area
full-service label system
business scene portrait
multi-dimensional category features
