基于规则推理与自然语言处理的停复电调控行为防误技术
CSTR:
作者:
作者单位:

(1.广西电网有限责任公司电力调度控制中心,南宁 530000;2.南京南瑞信息通信科技有限公司,南京 211100)

作者简介:

傅源(1994-),男,广西壮族自治区北海市人,工程师,主要研究方向为电网调度运行技术等。

通讯作者:

中图分类号:

TM76;TP391

基金项目:

调控行为辅助防误及评价关键技术研究(046000KK52222030).


Error-Prevention Techniques for Power On/Off Regulation Behavior Based on Rule-Based Reasoning and Natural Language Processing
Author:
Affiliation:

(1. Guangxi Master Electric Power Dispatching and Control Center Guangxi Power Co.,Ltd.,Nanning 530000,CHN;2. Nanjing NARI Information and Communication Technology Co.,Ltd.,Nanjing 211100,CHN)

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    文章提出一种基于规则推理与自然语言处理的停复电调控行为防误技术。该方法首先利用 iStar模型构建调控行为业务数据模型,在整合多源异构数据的基础上,通过精化关系、 NeededBy关系、Qualification关系表征调控行为与停复电事件之间的关联。然后,对上述关系中的检修单自然语言信息进行词嵌入与语句向量化处理,并采用 LCS算法提取同类样本的公共词语,形成停复电调控行为的防误规则。最后,在 EBRB推理机制下对防误规则进行推理,获得规则的激活概率,从而实现防误功能。研究结果表明,在防误操作阶段,操作票执行合格率稳定在 95%以上,工作票的正确执行次数显著优于对照组,所提方法能够有效提升停复电调控操作的安全性。

    Abstract:

    This paper proposes an error-prevention technique for power on/off regulation behavior based on rule-based reasoning and natural language processing. First,the iStar model is employed to construct a business data model for regulation behavior. After integrating multisource heterogeneous data, the relationships between regulation behavior and power on/off events are characterized through refinement, NeededBy, and Qualification relationships. Second, word embedding and sentence vectorization are applied to the natural language information of maintenance tickets within these relationships,and the LCS algorithm is used to extract common phrases from similar samples to form error-prevention rules for power on/off regulation behavior. Finally,the EBRB inference mechanism reasons over these error-prevention rules to obtain the activation probability of each rule,thereby achieving error prevention. The results show that during the error-prevention phase,the qualified rate of operation-ticket execution remains above 95%,and the number of correctly executed work orders is significantly higher than that of the control group. The proposed method effectively enhances the safety of power on/off regulation operations.

    参考文献
    相似文献
    引证文献
引用本文

傅源,汪成军,赵嘉伦,阳晟,刘津铭.基于规则推理与自然语言处理的停复电调控行为防误技术[J].半导体光电,2026,47(3):575-580. FU Yuan, WANG Chengjun, ZHAO Jialun, YANG Sheng, LIU Jinming. Error-Prevention Techniques for Power On/Off Regulation Behavior Based on Rule-Based Reasoning and Natural Language Processing[J].,2026,47(3):575-580.

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2025-08-14
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2026-06-25
  • 出版日期:
文章二维码