报告详情
  1. 活动首页
  2. 报告清单
  3. 报告详情
Optimization of Substation Monitoring Message Table Based on Word Embedding
编号：68 访问权限：仅限参会人 更新：2026-09-21 22:58:59 浏览：13次 口头报告

报告开始：暂无开始时间（Asia/Shanghai）

报告时间：暂无持续时间

所在会场：[暂无会议] [暂无会议段]

暂无文件

摘要
To address the challenges of vendor-specific expression variations in monitoring information messages, the time-consuming nature of manual item-by-item organization, and the over-segmentation of compound terminology during the integrated automation upgrade of substations, this paper proposes an intelligent optimization method for monitoring information messages based on word embedding techniques. First, the messages are structurally segmented into header, body, and footer sections, and a synonym library, a segmentation dictionary, and a standard term library are established. Second, the Skip-Gram model is adopted to construct professional terminology vectors, and cosine similarity is employed to achieve semantic matching between non-standard and standard terms. Furthermore, a "magnetic-attraction-inspired" defragmentation mechanism tailored for compound electrical terminology is introduced, leveraging business attributes — including substation, voltage level, associated bay, and equipment type — to constrain the candidate set. To mitigate the risk of erroneous automatic output, a decision rule incorporating both a similarity threshold and a candidate gap threshold is formulated, and parameter optimization is conducted with the F1 score, manual review rate, and processing time as objective functions. Simulation validation is performed on 12,000 monitoring messages. The results demonstrate that the proposed method achieves an F1 score of 96.7% and reduces the manual review rate to 6.4%. Moreover, the average processing time for monitoring information tables is reduced by 68.72%, yielding an approximately 3.20-fold improvement in overall processing efficiency. These results confirm that the proposed method can significantly reduce the manual effort required for organizing monitoring information messages while preserving readability.
关键词
Word embedding, Skip-Gram, Semantic similarity, Fuzzy matching, Monitoring information message
报告人
Yichao Wang
Engineer Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd.

稿件作者
Yichao Wang Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd.
Lu Sun Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd
Yanbo Zhao Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd
Meng Sun Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd.
Hui Wan Hohhot Power Supply Branch Company Inner Mongolia Power (Group) Co., Ltd.
Zirui Wang Hohhot Power Supply Company
发表评论
验证码 看不清楚，更换一张
全部评论
重要日期

  • 会议日期

    11月06日

    2026

    至

    11月08日

    2026

  • 10月15日 2026

    初稿截稿日期

主办单位
IEEE Instrumentation and Measurement Society
承办单位
Sichuan University
联系方式
登录查看完整联系方式
历届会议
移动端
在手机上打开
小程序
打开微信小程序
客服
扫码或点此咨询