Si Jiandong, Liu Chang, Ye Jingxian, Wu Jianfeng, Wang Jianguo, Hu Kairui, Ju Chunhua, Cao Qianwen
State Grid Jinhua Power Supply Company, Jinhua, Zhejiang, China.
Modern Business Research Center, Zhejiang Gongshang University, Hangzhou, Zhejiang, China.
Front Big Data. 2025 Jun 6;8:1603106. doi: 10.3389/fdata.2025.1603106. eCollection 2025.
The decision regarding the supply of emergency equipments for power emergencies requires timeliness, efficiency, and accuracy. The multi-agent supply relationship graph, based on complex data fusion, enables the comprehensive exploration of interconnections among key entities in power emergency supplies.
This approach enhances decision-making efficiency and quality by uncovering multiple relationships between main bodies involved. The present study focuses on the decision-making process for power emergency equipments supply and aims to enhance its professionalization. To achieve this goal, multi-modal data regarding power emergency equipments supply is collected from both internal and external power enterprises. Subsequently, a decision support knowledge base is established, along with a four-dimensional relationship graph that integrates events, time, equipments, and suppliers based on the knowledge graph. This enables the mining of multidimensional relationships pertaining to the main body. Finally, supported by the graph, the platform can offer intelligent assistance in decision-making, supplier recommendation, optimization of emergency equipment scheduling for electric power supply, and provides effective information and guidance for decision-making in electric power emergency equipment supply.
After conducting a comparative analysis, the decision support system based on the knowledge graph proposed in this study demonstrates superior effectiveness and precision. By integrating the four-dimensional relationship graph with data mining algorithms, precise decision support can be provided for power emergency response. After verification through case studies, the model developed in this study was utilized to recommend suppliers of power emergency equipment, and the recommendation results demonstrated a closer alignment with actual procurement outcomes.
This system proposed by this study delivers multidimensional knowledge guidance and optimized decision pathways for emergency supply management.
关于电力紧急情况应急设备供应的决策需要及时性、效率和准确性。基于复杂数据融合的多智能体供应关系图能够全面探索电力应急供应中关键实体之间的相互联系。
这种方法通过揭示相关主体之间的多重关系来提高决策效率和质量。本研究聚焦于电力应急设备供应的决策过程，旨在提升其专业化水平。为实现这一目标，从电力企业内部和外部收集了有关电力应急设备供应的多模态数据。随后，建立了一个决策支持知识库，以及一个基于知识图谱整合事件、时间、设备和供应商的四维关系图。这使得能够挖掘与主体相关的多维关系。最后，在该图的支持下，该平台可为决策提供智能辅助、供应商推荐、优化电力供应应急设备调度，并为电力应急设备供应决策提供有效的信息和指导。
经过对比分析，本研究提出的基于知识图谱的决策支持系统展现出卓越的有效性和精确性。通过将四维关系图与数据挖掘算法相结合，可为电力应急响应提供精确的决策支持。经案例研究验证，本研究开发的模型被用于推荐电力应急设备供应商，推荐结果与实际采购结果更为契合。
本研究提出的该系统为应急供应管理提供了多维知识指导和优化的决策路径。
