一种基于多模态信息的输变电线路故障识别方法

By constructing millisecond-level synchronous multimodal data acquisition and dynamic attention weight allocation and cross-modal semantic alignment technology, combined with graph neural networks, the spatiotemporal alignment and feature fusion problems of multimodal data in power transmission and transformation line fault identification are solved, achieving high-precision and robust fault identification, and improving identification accuracy and response speed.

CN121458938BActive Publication Date: 2026-07-17BEIJING ANXIN YIWEI TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BEIJING ANXIN YIWEI TECH CO LTD
Filing Date
2025-11-05
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing technologies struggle to achieve spatiotemporal alignment of multimodal data, feature complementarity mining, and dynamic weight allocation in transmission and transformation line fault identification. This leads to weakened semantic associations between modes, an inability to effectively address data gaps, noise interference, or mode asynchrony, and a lack of prior modeling capabilities for the evolution of typical transmission and transformation line faults, resulting in insufficient identification accuracy and response speed.

Method used

Employing a millisecond-level synchronous multimodal data acquisition mechanism, and combining dynamic attention weight allocation and cross-modal semantic alignment techniques with graph neural networks to embed the evolution patterns of typical faults in power transmission and transformation lines, a multimodal feature fusion network is constructed to achieve accurate identification of fault type, location, and severity, and improve robustness under low signal-to-noise ratio conditions.

Benefits of technology

It significantly improves the accuracy and real-time performance of fault identification, reduces false alarm and false negative rates, has strong generalization capabilities, and provides support for the safe and stable operation of smart grids.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121458938B_ABST
    Figure CN121458938B_ABST
Patent Text Reader

Abstract

本发明涉及电力系统输配电技术领域，且公开了一种基于多模态信息的输变电线路故障识别方法，旨在解决现有技术中因模态割裂、同步性差及模型缺乏物理先验导致的误报率高、鲁棒性不足的问题。该方法包括：同步采集电气量、环境参数、可见光图像、红外热成像及声学信号等多模态数据并实现毫秒级时间对齐；分别提取各模态的暂态、结构、热分布及频域特征；通过模态间注意力权重分配与跨模态语义对齐模块进行动态特征融合；将融合特征输入嵌入线路故障演化规律的图神经网络进行故障判别；输出故障类型、位置、严重等级及置信度，通过上述技术方案，实现了高精度、强鲁棒、低延迟的多模态协同诊断，显著提升了输变电线路故障识别的准确性与可靠性。
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Line holographic anomaly detection method and system based on cross-modal intelligent collaboration

    CN120579035A

  • Micro-grid fault diagnosis and dynamic recovery method based on deep reinforcement learning

    CN120597036A