面向电网调度领域的实体识别技术

doi:10.12204/j.issn.1000-7229.2021.10.008

电力建设 ›› 2021, Vol. 42 ›› Issue (10): 71-77.doi: 10.12204/j.issn.1000-7229.2021.10.008

面向电网调度领域的实体识别技术

徐会芳(), 张中浩, 谈元鹏, 韩富佳

中国电力科学研究院有限公司,北京市100192

收稿日期:2020-11-30 出版日期:2021-10-01 发布日期:2021-09-30
通讯作者: 徐会芳 E-mail:xuhuifang@epri.sgcc.com.cn
作者简介:张中浩(1991),男,博士,工程师,研究方向为电网智能运维、知识图谱等。
谈元鹏(1987),男,博士,高级工程师,研究方向为电力系统自动化、知识图谱、目标检测识别等。
韩富佳(1989),男,博士,工程师,研究方向为智能电网大数据分析、人工智能应用等。
基金资助:
国家电网有限公司科技项目“知识图谱在电网故障处理中的应用关键技术研究”(SGJB0000TKJS1900099)

Research on Entity Recognition Technology in Power Grid Dispatching Field

XU Huifang(), ZHANG Zhonghao, TAN Yuanpeng, HAN Fujia

China Electric Power Research Institute, Beijing 100192, China

Received:2020-11-30 Online:2021-10-01 Published:2021-09-30
Contact: XU Huifang E-mail:xuhuifang@epri.sgcc.com.cn

摘要/Abstract

摘要：

近年随着电网调度领域数据自动化、智能化管理需求的日益增长,知识图谱成为提供知识管理、智能查询、辅助决策等功能的重要技术。实体作为构成知识图谱的核心要素,识别的准确率将直接影响知识图谱的质量。针对电网调度领域,首先分析电网调度实体识别研究现状,明确了实体识别任务目标,然后根据电网调度领域文本数据特征,设计了同时满足局部特征与全局特征提取需求的算法结构,构建了基于双向长短期记忆网络(bi-directional long short-term memory, BiLSTM)-卷积神经网络(convolutional neural networks, CNN)-条件随机场(conditional random field,CRF)的电网调度领域实体识别模型。最后实验结果表明,所提方法识别准确率达到93.1%,F₁值达到86.05%,能够有效支撑电网调度领域实体识别工作的开展。

关键词: 实体识别, 知识图谱, 双向长短期记忆网络(BiLSTM), 卷积神经网络(CNN), 条件随机场(CRF)

Abstract:

In recent years, with the increasing demand of data automation and intelligent management in the field of power grid dispatching, knowledge graph has become an important technology to provide knowledge management, intelligent query, auxiliary decision-making and other functions. As the core element of knowledge graph, the accuracy of entity recognition will directly affect the quality of knowledge graph. Aiming at the field of power grid dispatching, this paper firstly analyzes the research status of entity recognition in power grid dispatching field, and defines the task objective of entity recognition. Then, according to the text data features of power grid dispatching, an algorithm structure is designed to meet the requirements of local and global feature extraction, and a named entity recognition model based on BiLSTM-CNN-CRF is constructed. Finally, the experimental results show that the recognition accuracy of this method reaches 93.1%, and the F₁ value reaches 86.05%, which can effectively support the development of entity recognition in the field of power grid dispatching.

Key words: entity recognition, knowledge graph, bi-directional long short-term memory (BiLSTM), convolutional neural networks (CNN), conditional random field(CRF)

中图分类号:

TM734

徐会芳, 张中浩, 谈元鹏, 韩富佳. 面向电网调度领域的实体识别技术[J]. 电力建设, 2021, 42(10): 71-77.

XU Huifang, ZHANG Zhonghao, TAN Yuanpeng, HAN Fujia. Research on Entity Recognition Technology in Power Grid Dispatching Field[J]. ELECTRIC POWER CONSTRUCTION, 2021, 42(10): 71-77.

图/表 9

图1 BiLSTM模型

Fig.1 BiLSTM model

图2 BiLSTM-CNN-CRF模型

Fig.2 BiLSTM-CNN-CRF model

表1 实体识别示例

Table 1 Examples of entity recognition

图3 不同模型训练中的损失函数值

Fig.3 Loss value in different model training

表2 Softmax分类器实体识别结果

Table 2 Entity recognition results of Softmax classifier %

表3 CRF分类器实体识别结果

Table 3 Entity recognition results of CRF classifier %

图4 分类器对应的F1值

Fig.4 The corresponding F1 of classifier

图5 最优模型结果对比

Fig.5 Result comparison of optimal models

表4 实体识别结果

Table 4 Entity recognition results %

参考文献 20

[1]	李新鹏, 徐建航, 郭子明, 等. 调度自动化系统知识图谱的构建与应用[J]. 中国电力, 2019, 52(2): 70-77,157.
	LI Xinpeng, XU Jianhang, GUO Ziming, et al. Construction and application of knowledge graph of power dispatching automation system[J]. Electric Power, 2019, 52(2): 70-77,157.
[2]	周帆, 叶健辉, 肖林朋, 等. 基于知识图谱的电网模型本体智能问答系统研究[J]. 中国科技信息, 2019(16): 85-86.
[3]	闪鑫, 陆晓, 翟明玉, 等. 人工智能应用于电网调控的关键技术分析[J]. 电力系统自动化, 2019, 43(1): 49-57.
	SHAN Xin, LU Xiao, ZHAI Mingyu, et al. Analysis of key technologies for artificial intelligence applied to power grid dispatch and control[J]. Automation of Electric Power Systems, 2019, 43(1): 49-57.
[4]	刘梓权, 王慧芳. 基于知识图谱技术的电力设备缺陷记录检索方法[J]. 电力系统自动化, 2018, 42(14): 158-164.
	LIU Ziquan, WANG Huifang. Retrieval method for defect records of power equipment based on knowledge graph technology[J]. Automation of Electric Power Systems, 2018, 42(14): 158-164.
[5]	ZHOU P, SHI W, TIAN J, et al. Attention-based bidirectional long short-term memory networks for relation classification[C]// Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA, USA: Association for Computational Linguistics, 2016: 207-212.
[6]	XU K, FENG Y S, HUANG S F, et al. Semantic relation classification via convolutional neural networks with simple negative sampling[C]// Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing. Stroudsburg, PA, USA: Association for Computational Linguistics, 2015: 536-540.
[7]	吴超. 电力调度知识图谱中知识抽取系统的设计与实现[D]. 沈阳: 中国科学院大学(中国科学院沈阳计算技术研究所), 2020.
	WU Chao. Design and implementation of knowledge extraction system in knowledge graph of electric power dispatch[D]. Shenyang: Shengyang Institute of Computing Technology, University of Chinese Academy of Sciences, 2020.
[8]	王适乾. 电力调度控制系统中语义解析技术研究[D]. 济南: 山东大学, 2018.
	WANG Shiqian. Research on semantic parsing technology in power control system[D]. Jinan: Shandong University, 2018.
[9]	LIU Y, WEI F R, LI S J, et al. A dependency-based neural network for relation classification[C]// Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing. Stroudsburg, PA, USA: Association for Computational Linguistics, 2015.
[10]	REI M, CRICHTON G, PYYSALO S. Attending to characters in neural sequence labeling models[C]// Proceedings of the 26th International Conference on Computational Linguistics. Osaka,Japan: ICCL, 2016:309-318.
[11]	王莉军, 周越, 桂婕, 等. 基于BiLSTM-CRF的中医文言文文献分词模型研究[J]. 计算机应用研究, 2020, 37(11): 3359-3362,3367.
	WANG Lijun, ZHOU Yue, GUI Jie, et al. Chinese medicine literature word segmentation based on BiLSTM-CRF[J]. Application Research of Computers, 2020, 37(11): 3359-3362,3367.
[12]	JI Z X, WANG X H, CAI C Y, et al. Power entity recognition based on bidirectional long short-term memory and conditional random fields[J]. Global Energy Interconnection, 2020, 3(2): 186-192. doi: 10.1016/j.gloei.2020.05.010 URL
[13]	刘宇鹏, 栗冬冬. 基于BLSTM-CNN-CRF的中文命名实体识别方法[J]. 哈尔滨理工大学学报, 2020, 25(1): 115-120.
	LIU Yupeng, LI Dongdong. Chinese named entity recognition method based on Bi-directional LSTM-CNN-CRF[J]. Journal of Harbin University of Science and Technology, 2020, 25(1): 115-120.
[14]	高翔, 张金登, 许潇, 等. 基于LSTM-CRF的军事动向文本实体识别方法[J]. 指挥信息系统与技术, 2020, 11(6): 91-95.
	GAO Xiang, ZHANG Jindeng, XU Xiao, et al. Military trend text entity recognition method based on LSTM-CRF[J]. Command Information System and Technology, 2020, 11(6): 91-95.
[15]	郑诚, 洪彤彤, 薛满意. 用于短文本分类的BLSTM_MLPCNN模型[J]. 计算机科学, 2019, 46(6): 206-211.
	ZHENG Cheng, HONG Tongtong, XUE Manyi. BLSTM_MLPCNN model for short text classification[J]. Computer Science, 2019, 46(6): 206-211.
[16]	REKIA KADARI. 基于深度学习模型的CCG超标注[D]. 哈尔滨: 哈尔滨工业大学, 2018.
	REKIA KADARI. CCG supertagging based on deep learning models[D]. Harbin: Harbin Institute of Technology, 2018.
[17]	AN C, HUANG J M, CHANG S F, et al. Question similarity modeling with bidirectional long short-term memory neural network[C]// 2016 IEEE First International Conference on Data Science in Cyberspace (DSC). Changsha, China: IEEE, 2016: 318-322.
[18]	RAY A, RAJESWAR S, CHAUDHURY S. Text recognition using deep BLSTM networks[C]// 2015 Eighth International Conference on Advances in Pattern Recognition (ICAPR). Kolkata, India: IEEE, 2015: 1-6.
[19]	石民, 李斌, 陈小荷. 基于CRF的先秦汉语分词标注一体化研究[J]. 中文信息学报, 2010, 24(2): 39-45.
	SHI Min, LI Bin, CHEN Xiaohe. CRF based research on a unified approach to word segmentation and POS tagging for pre-Qin Chinese[J]. Journal of Chinese Information Processing, 2010, 24(2): 39-45.
[20]	张金龙, 王石, 钱存发. 基于CRF和规则的中文医疗机构名称识别[J]. 计算机应用与软件, 2014, 31(3): 159-162,198.
	ZHANG Jinlong, WANG Shi, QIAN Cunfa. Crf and rules-based recognition of medical institutions name in Chinese[J]. Computer Applications and Software, 2014, 31(3): 159-162,198.

面向电网调度领域的实体识别技术

Research on Entity Recognition Technology in Power Grid Dispatching Field

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 20

相关文章 3

编辑推荐

Metrics

本文评价

[1]	肖发龙, 吴岳忠, 沈雪豪, 何震凯, 秦烨. 基于深度学习和知识图谱的变电站设备故障智能诊断[J]. 电力建设, 2022, 43(3): 66-74.
[2]	庄家懿, 杨国华, 郑豪丰, 王煜东, 胡瑞琨, 丁旭. 并行多模型融合的混合神经网络超短期负荷预测[J]. 电力建设, 2020, 41(10): 1-8.
[3]	牛东晓，马天男，王海潮，刘鸿飞，黄雅莉. 基于KPCA和NSGAII优化CNN参数的电动汽车充电站短期负荷预测[J]. 电力建设, 2017, 38(3): 85-.