摘要
针对多源异构分布数据存储混乱、融合计算性能不足等问题,提出了基于边缘计算的智能配电网多源数据处理与融合方法。首先,设计了基于边缘计算的配电物联网多源数据处理与融合体系结构;然后,提出基于改进门控循环单元(gate recurrent unit,GRU)网络的配电网多源数据处理与融合模型,引入一种改进的GRU进行数据融合;并在此基础上,增加了注意力机制,减少历史数据的丢失,增强重要信息的影响力,从而实现数据的精准处理与融合。通过试验分析,发现所提模型对低压配电网多源数据的处理效率更高、计算速度更快、误差更小。
Aiming at the problems such as multi-source heterogeneous distributed data storage confusion and insufficient fusion computing performance,a multi-source data processing and fusion method for intelligent distribution network based on edge computing was proposed.Firstly,the multi-source data processing and fusion architecture of distribution IoT based on edge computing was designed;then,a multi-source data processing and fusion model for distribution networks based on an improved gate recurrent unit(GRU)network was proposed,and an improved GRU was introduced for data fusion;on this basis,the attention mechanism was added to reduce the loss of historical data,enhance the influence of important information,and achieve precise data processing and fusion.Through experimental analysis,it is found that the proposed model has higher processing efficiency,faster calculation speed,and smaller errors for multi-source data in low-voltage distribution networks.
作者
赵永柱
Zhao Yongzhu(State Grid Shaanxi Electric Power Co.,Ltd.,Xi’an Shaanxi 710048,China)
出处
《电气自动化》
2024年第5期79-81,共3页
Electrical Automation
关键词
边缘计算
配电网
门控循环单元(GRU)
多源数据
融合
注意力机制
edge calculation
distribution network
gate recurrent unit(GRU)
multi-source data
fusion
attention mechanism
作者简介
赵永柱(1974-),男,陕西咸阳人,硕士,高级工程师,研究方向:计算机应用和电力信息化。
