A terminal electric carbon data anomaly detection method based on dynamic fusion

By employing a dynamic modal gating fusion mechanism and edge-cloud collaborative optimization, the problems of root cause localization difficulties and insufficient model generalization ability in the detection of carbon anomalies in electricity have been solved, thereby improving the interpretability and adaptability of carbon monitoring.

CN122413253APending Publication Date: 2026-07-17NINGXIA LGG INSTR CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610859728.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-06-15
Publication Date
2026-07-17

Smart Images

  • Figure CN122413253A_ABST
    Figure CN122413253A_ABST
Patent Text Reader

Abstract

本发明涉及智能电网与边缘计算技术领域，公开了一种基于动态融合的终端电碳数据异常检测方法，解决现有技术中电碳异常检测结果缺乏可解释性，难以定位故障根源、边缘侧模型训练与更新机制孤立，跨台区泛化能力不足的问题。本发明利用动态融合权重偏差，将异常自动映射为碳因子波动、负荷突变或环境因素，为运维提供直观告警依据，克服黑盒模型难以信任的缺陷；基于历史残差标准差与均值构建前瞻校准因子，动态修正负荷预测置信区间，结合连续越限判定，显著降低非平稳工况下的误报率；云端对同组终端碳排放模型参数进行联邦平均并下发，实现跨台区知识共享，提升模型对拓扑变化的适应性与收敛速度。
Need to check novelty before this filing date? Find Prior Art