An online early warning and self-healing control method for power grid faults based on an agent

By deploying intelligent agents in the power grid, real-time collection of electrical quantity data and dynamic health index calculation and risk map generation solve the problems of high false alarm rate and insufficient dynamic assessment capability of existing power grid fault early warning methods, realizing high-precision fault early warning and self-healing control, and improving the resilience and intelligence level of the distribution network.

CN122371445APending Publication Date: 2026-07-10SICHUAN RES INST OF SHANGHAI JIAOTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610249799.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-03
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing power grid fault early warning methods cannot effectively distinguish between normal disturbances caused by new energy fluctuations and random load changes and actual fault precursors, resulting in high false alarm and false alarm rates. They also lack the ability to dynamically assess the propagation path and impact range of faults in the power grid topology, making it difficult to achieve regional-level early risk perception and proactive intervention.

Method used

Deploy intelligent agents for monitoring, early warning, and regulation. Calculate dynamic health index by collecting electrical quantity data in real time, generate a fault propagation risk map by combining graph attention network, and use spatiotemporal decay weights for risk weighted fusion to construct a rolling optimization self-healing regulation objective function, generate the optimal self-healing control strategy, and conduct safety verification by combining a digital twin model.

Benefits of technology

It achieves high-precision fault early warning and self-healing control, reduces false alarm rate, improves the accuracy and timeliness of fault information, ensures the strong robustness of self-healing strategy in uncertain environments, and enhances the resilience and intelligence level of distribution network.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122371445A_ABST
    Figure CN122371445A_ABST
Patent Text Reader

Abstract

This invention discloses an agent-based online early warning and self-healing control method for power grid faults, relating to the field of smart distribution network technology. It includes deploying monitoring agents, early warning agents, and control agents in the distribution network. The monitoring agent collects voltage, current, and power electrical quantity data of its local node in real time and sends the collected electrical quantity data along with timestamps to the corresponding early warning agent. The early warning agent calculates an original dynamic health index sequence from the received electrical quantity data and performs modal decomposition on the original dynamic health index sequence to extract low-frequency components characterizing long-term operating trends. These low-frequency components are used as a modified dynamic health index to determine whether an early warning is triggered. When the modified dynamic health index is lower than a preset threshold, the early warning agent combines the rate of change of the modified dynamic health index of neighboring nodes with the electrical distance to generate a fault propagation risk map with timestamps through a graph attention network.
Need to check novelty before this filing date? Find Prior Art