面向智能巡检终端的非结构化数据特征提取技术  

Feature extraction technology of unstructured data for intelligent inspection terminal

作　　者：罗劲斌 章坚 郭启迪 李端姣 李雄刚 LUO Jinbin;ZHANG Jian;GUO Qidi;LI Duanjiao;LI Xionggang(Machine Patrol Management Center of Guangdong Power Grid Co.,Ltd.,Guangzhou 510335,China)

机构地区：[1]广东电网有限责任公司机巡管理中心,广东广州510335

出　　处：《电子设计工程》2025年第1期100-103,108,共5页Electronic Design Engineering

基　　金：南方电网公司科技项目资助(GDKJXM20210192)。

摘　　要：智能巡检终端采集的电力设备数据大多为图像、视频、声音等非结构化数据,具有复杂性、多样性的特征。对上述非结构化数据提取的准确性决定了电力设备的监测能力,为此,面向智能巡检终端中的非结构化数据,提出了一种新的特征提取技术。分别识别智能巡检终端中数据的图像特征值、视频特征值、声音特征值。以识别结果为基础,对其进行归一化处理,利用K-L变换完成对数据样本的降维处理,实现对智能巡检终端非结构化数据特征的提取。实验结果表明,所提方法提取的结构化数据样本长度始终与智能巡检终端主机所需输配电数据样本长度差距小于0.05×109 MB,提高了非结构化数据特征提取的精准性。Most of the power equipment data collected by intelligent inspection terminals are unstructured data such as images,videos and sounds,which are characterized by complexity and diversity.The accuracy of unstructured data extraction determines the monitoring capability of power equipment.Therefore,a new feature extraction technology is proposed for unstructured data in intelligent patrol terminal.Identify the image feature values,video feature values,and sound feature values of the data in the intelligent inspection terminal separately.Based on the recognition results,after normalization,K-L transformation is used to complete the dimensionality reduction of data samples,and the feature extraction of unstructured data of intelligent patrol terminal is realized.The experimental results show that the difference between the sample length of structured data extracted by the proposed method and the sample length of transmission and distribution data required by intelligent patrol terminal host is always less than 0.05×109 MB,improving the accuracy of unstructured data feature extraction.

关 键 词：智能巡检终端 非结构化数据 特征提取 K-L变换 数据降维 

分 类 号：TN02[电子电信—物理电子学]