• CSCD核心库收录期刊
  • 中文核心期刊
  • 中国科技核心期刊

电力建设 ›› 2020, Vol. 41 ›› Issue (7): 17-24.doi: 10.12204/j.issn.1000-7229.2020.07.003

• 能源互联网环境下电力系统安全可靠、经济运行 ·栏目主持 黄涛研究员· • 上一篇    下一篇

基于运行关键指标和Seq2Seq的大电网运行异常识别

庞传军1,2,牟佳男3,余建明1,2,武力3   

  1. 1. 南瑞集团有限公司(国网电力科学研究院有限公司), 南京市 211106;2. 北京科东电力控制系统有限责任公司,北京市 100192;3. 国家电网有限公司国家电力调度控制中心,北京市 100031
  • 出版日期:2020-07-01
  • 作者简介:庞传军(1984),男,硕士,高级工程师,主要研究方向为电力系统自动化、机器学习技术在电力系统中的应用; 牟佳男(1987),男,博士,高级工程师,主要研究方向为电网调度运行与控制; 余建明(1979),男,博士,高级工程师,主要研究方向为电力系统自动化、机器学习技术在电力系统中的应用; 武力(1985),男,硕士,高级工程师,主要研究方向为电网调度运行与控制、电力系统及其自动化。
  • 基金资助:
    国家电网公司科技项目(5100-201940013A-0-0-00)

Identification of Abnormal Operation of Large Power Grids According to Key Operating Indicators and Seq2Seq

PANG Chuanjun1,2,MOU Jianan 3,YU Jianming1,2,WU Li3   

  1. 1.NARI Group Corporation Co., Ltd. (State Grid Electric Power Research Institute Co., Ltd.), Nanjing 211106, China; 2.Beijing Kedong Electric Power Control System Co., Ltd., Beijing 100192, China; 3.Power Dispatching and Control Center ,State Grid Corporation of China, Beijing 100031, China
  • Online:2020-07-01
  • Supported by:
    This work is supported by State Grid Corporation of China Research Program(No. 5100-201940013A-0-0-00).

摘要: 基于电网运行指标的态势感知是未来主要电网调度模式,电网运行异常识别是态势感知的重要内容。首先,从大电网调控运行需求出发构建了全面反映电网运行情况的综合指标体系;然后,采用由长短期记忆单元组成的自动编码机构建指标异常识别模型,在缺少电网运行指标异常数据的情况下,采用无监督的方式从电网正常运行状态下指标历史数据中学习指标的内在模式;最后,基于模型的重构误差分布提出了反映指标偏离正常状态的异常分数。将电网运行指标实时数据送入训练后的模型进行重构,当存在异常时会产生较大的异常分数,根据异常分数识别电网运行指标异常。实验结果表明,当电网运行指标出现异常时该模型可以根据异常分数进行有效识别,从而帮助电网调度人员及时感知电网运行风险,及时采取控制措施保障电网运行安全。

关键词: 运行指标, 异常识别, 序列到序列(Seq2Seq), 长短期记忆单元(LSTM), 电网运行

Abstract: Situation awareness based on power grid operation indicators is the trend of future dispatching mode. Identification of anomalies is an important content of situation awareness. A comprehensive indicator system that comprehensively reflects the power grid operation is constructed. An auto-encoder composed of LSTM is used to construct the index abnormality identification model. In the absence of abnormal data on power grid operation indicators, an unsupervised approach is adopted to learn the internal model of the indicators from the historical data of the indicators under normal operating conditions of the power grid. On the basis of model reconstruction error distribution, this paper proposes an abnormal score reflecting the deviation of the index from the normal state. The real-time data of the grid operation indicators are sent to the trained model for reconstruction, and a large abnormal score will be generated when there is an abnormality. The experimental results show that the model can effectively identify the abnormal scores when the grid operation indicators are abnormal. This helps grid dispatchers perceive grid operation risks in a timely manner and take timely control measures to ensure grid security.

Key words: operating indicators, anomaly identification, sequence to sequence(Seq2Seq), long short term memory(LSTM), power grid operation

中图分类号: