Scholarlink

<p>This paper proposes an integrated knowledge graph-based power dispatching model for emergency response, combining Markov chain-based text preprocessing, entity-extracted knowledge graph construction, and case-based reasoning optimization - a novel approach that enhances both real-time decision-making and system security. First, a Markov chain-based method effectively removes redundant information from power anomaly event texts, improving entity extraction accuracy. Subsequently, a knowledge graph is constructed to precisely identify key entities, enabling the creation of a structured power emergency plan database. Finally, case-based reasoning matches real-time anomalies with historical cases, facilitating the rapid generation of optimal dispatching schemes. The experiments demonstrate that the proposed model achieves high efficiency (with an average dispatching time &lt; 50&#xa0;s) and reliability (exhibiting a failure blowout rate below 0.1%), thereby significantly improving power grid safety. The proposed framework advances intelligent power system dispatching by integrating text analytics, knowledge representation, and adaptive reasoning.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on power dispatching model based on knowledge graph entity extraction task

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月18日

查看全文
  • Yufeng Chai,
  • Bo Zhang,
  • Min Wang,
  • Zhongying Zhao

摘要

This paper proposes an integrated knowledge graph-based power dispatching model for emergency response, combining Markov chain-based text preprocessing, entity-extracted knowledge graph construction, and case-based reasoning optimization - a novel approach that enhances both real-time decision-making and system security. First, a Markov chain-based method effectively removes redundant information from power anomaly event texts, improving entity extraction accuracy. Subsequently, a knowledge graph is constructed to precisely identify key entities, enabling the creation of a structured power emergency plan database. Finally, case-based reasoning matches real-time anomalies with historical cases, facilitating the rapid generation of optimal dispatching schemes. The experiments demonstrate that the proposed model achieves high efficiency (with an average dispatching time < 50 s) and reliability (exhibiting a failure blowout rate below 0.1%), thereby significantly improving power grid safety. The proposed framework advances intelligent power system dispatching by integrating text analytics, knowledge representation, and adaptive reasoning.