基于深度学习的智能录波器配置数据自动化映射方法

李铁成; 任江波; 刘清泉; 耿少博; 王志华; 周达明

李铁成,任江波,刘清泉,耿少博,王志华,周达明.基于深度学习的智能录波器配置数据自动化映射方法[J].电测与仪表,2022,59(9):76-83.
LI Tiecheng,6,3,4,5,ZHOU Daming.Automatic Mapping Method of Intelligent Recorder Configuration Datasets based on Deep Learning[J].Electrical Measurement & Instrumentation,2022,59(9):76-83.

基于深度学习的智能录波器配置数据自动化映射方法

Automatic Mapping Method of Intelligent Recorder Configuration Datasets based on Deep Learning

DOI：10.19753/j.issn1001-1390.2002.09.011

中文关键词: 智能变电站文本挖掘全站配置描述文件智能二次设备动态卷积神经网络

英文关键词:intelligent substation text mining Substation Configuration Description Intelligent Electronic Device Dynamic Convolutional Neural Network

基金项目:

作者中文名	作者英文名	单位
李铁成	LI Tiecheng	国网河北省电力有限公司电力科学研究院
任江波	6	6
刘清泉	3	3
耿少博	4	4
王志华	5	5
周达明	ZHOU Daming	武汉凯默电气有限公司

摘要点击次数: 1986

中文摘要:

智能变电站配置描述文件中包含大量智能二次设备数据输出接口地址的配置数据集，将这些数据集映射至智能录波器各信息组是保证录波器精准采集设备运行数据的基础性步骤，当前主流映射方法是依照输出接口描述文本人工映射对应的配置数据，二次设备数目繁多时映射工作量大，而描述文本一定程度的不规范性给数据集自动化映射提出了难题。针对这一问题，本文提出了基于深度学习框架—动态卷积神经网络构造的智能录波器配置数据的自动化映射方法；首先利用文本表征模型word2vec对数据集描述文本的稀疏文本向量进行词组语义及关联关系的表征；随后构造动态卷积神经网络并输入文本向量，基于其多层次抽象化学习典型样本特征的特点进行语义规律挖掘与文本分类映射，据此结果实现接口地址配置数据的自动化映射。实际算例表明，基于动态卷积神经网络模型的文本分类方法语义分析能力强，分类精度高，有效提升了智能录波器配置数据自动化映射的准确率。

英文摘要:

The Substation Configuration Description (SCD) file of an intelligent substation contains a large number of Intelligent Electronic Devices (IEDs) configuration datasets of the output interface address. Mapping these datasets to the intelligent recorder is a basic step for the recorder to accurately collect IEDs’ operation information. The current mainstream mapping method is to map the corresponding configuration data manually according to the output interface description text. When the number of IEDs is extremely big, the configuration workload is large either, and since the text descriptions have a certain degree of irregularity, it also poses a problem for the automatic mapping of the datasets. Aiming at this problem, this paper proposes the automatic mapping technology of IEDs configuration datasets based on a deep learning framework—Dynamic Convolutional Neural Network (DCNN) model. Firstly, it uses the text representation model word2vec to represent the sparse text vectors of configuration datasets text descriptions as well as their phrase semantics and association relationships. Then text vectors will be imported in DCNN, which, based on its multilayer abstract learning characteristics of typical sample features, can perform semantic law mining and automatic mapping. The configuration datasets of intelligent recorder will be mapped automatically based on the text mapping result. The Practical example shows that the text classification method based on DCNN model has strong semantic analysis ability and high classification accuracy, which effectively improves the accuracy of automatic mapping of intelligent recorder configuration data.

查看全文查看/发表评论下载PDF阅读器

关闭