继电保护信息语义智能识别算法的研究与应用

钱海; 邱金辉; 张道农; 姜健琳; 尹浙洪; 沈亚东

引用本文:	钱海,邱金辉,张道农,等.继电保护信息语义智能识别算法的研究与应用[J].电力系统保护与控制,2018,46(3):83-88.
	QIAN Hai,QIU Jinhui,ZHANG Daonong,et al.Research and application of semantic intelligent recognition algorithm for relay protection information[J].Power System Protection and Control,2018,46(3):83-88

【打印本页】【下载PDF全文】【查看/发表评论】【EndNote】【RefMan】【BibTex】

←前一篇|后一篇→

过刊浏览高级检索

本文已被：浏览 6007次下载 2374次	码上扫一扫！
分享到：微信更多字体:加大+\|默认\|缩小-
继电保护信息语义智能识别算法的研究与应用
钱海¹,邱金辉¹,张道农²,姜健琳³,尹浙洪³,沈亚东³
(1.国网辽宁省电力有限公司，辽宁沈阳 110006;2.华北电力设计院有限公司，北京 100120; 3.上海泽鑫电力科技股份有限公司，上海 201206)

摘要:

为了应对当前存在于继电保护领域的语义不标准现象，设计一种继电保护信息语义智能识别算法。该智能算法基于字典管理、语义匹配、检索预处理以及检索管理4大模块。结合采集储存算法、预处理算法和文本相似度算法设计出一种智能自我学习的语义智能处理机制。该算法通过逐步提高计算机自识别率来实现对不标准继电保护语义的统计，从而丰富语义知识积累库，以便增加对不标准语义标准化的正确率。根据对不同类型不标准语义数据的测试，对得到的标准语义数据进行分析验证，证明了此继电保护信息语义智能识别算法具有准确的性能和良好的可行性。

关键词: 继电保护系统文本相似度算法人工智能

DOI：10.7667/PSPC170055

分类号:

基金项目:

Research and application of semantic intelligent recognition algorithm for relay protection information

QIAN Hai¹,QIU Jinhui¹,ZHANG Daonong²,JIANG Jianlin³,YIN Zhehong³,SHEN Yadong³

(1. State Grid Liaoning Electric Power Company Ltd, Shenyang 110006, China;2. North China Power Engineering Co., Ltd, Beijing 100120, China;3. Zexin Power Science & Technology Co., Ltd, Shanghai 201206, China)

Abstract:

In order to deal with the semantic non-standard phenomena in the field of relay protection currently, this paper designs a kind of relay protection information semantic intelligent recognition algorithm based on four modules including dictionary management, semantic matching, retrieval pre-processing, and retrieval management. Combined acquisition and storage algorithm, pre-processing algorithm, and text similarity algorithm, an intelligent self-learning based semantic intelligent processing mechanism is designed, which realizes the semantic statistics of non-standard relay protection through gradually improving computer self-recognition rate, thereby enriching semantic knowledge accumulation and then increasing the accuracy of non-standard semantic standardization. According to the testing of different types of non-standard semantic data, the obtained standard semantic datum are analyzed and verified, which proves that the relay protection information semantic intelligent recognition algorithm has good performance and feasibility.

Key words: relay protection system text similarity algorithm artificial intelligence

X关闭