一种基于融合终端的拓扑识别模型压缩方法、系统及设备

By pruning and compressing the topology recognition model and combining it with knowledge distillation, a formal model suitable for edge terminals is generated, solving the problem of deploying deep learning models on edge terminals and achieving efficient topology recognition and automatic updates.

CN119005263BActive Publication Date: 2026-07-17STATE GRID BEIJING ELECTRIC POWER CO +1
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
STATE GRID BEIJING ELECTRIC POWER CO
Filing Date
2024-08-15
Publication Date
2026-07-17

Smart Images

  • Figure CN119005263B_ABST
    Figure CN119005263B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于融合终端的拓扑识别模型压缩方法、系统及设备，属于数据处理技术领域，方法包括：获取融合终端采集的数据集；通过数据集，对初始拓扑识别模型进行训练，得到训练好的基础拓扑识别模型；对基础拓扑识别模型进行通道剪枝处理，得到剪枝处理后的拓扑识别模型；采用不同压缩方式，对剪枝处理后的拓扑识别模型进行压缩处理，得到不同版本的正式拓扑识别模型；根据融合终端的设备接入情况，向融合终端部署对应版本的正式拓扑识别模型。本发明可以根据设备接入情况向融合终端部署不同版本的拓扑识别模型，在模型精度不受到显著影响的前提下，减少计算资源消耗，提高模型在融合终端上运行的速度。
Need to check novelty before this filing date? Find Prior Art