基于边云智能协同的配电网信息物理系统

Distribution network information physics system based on edge cloud intelligent collaboration

随着配电网的发展,网络边缘设备产生的数据量和计算量急剧增加,带来了数据存储、云端计算和传输带宽压力。传统的云计算模式难以保证数据分析、处理与响应的实时性,致使信息流处理的非实时性直接影响到了能量流的稳定性与控制的可靠性。从边缘计算和云计算的协同应用角度出发,首先进行了边缘计算与云计算在配电网场景下的协同分析,然后构建了"自下而上,分布自治,协同调度,全局优化"的边云配电网信息物理系统模型,该模型能够更快地处理配电网的业务;最后,将深度学习应用在模型中实现配电网故障检查,验证了模型的高准确率。

With the development of distribution network,the amount of data and calculation generated by network edge equipment increases sharply,which brings the pressure of data storage,cloud computing and transmission bandwidth.Traditional cloud computing mode is difficult to ensure the realtime performance of data analysis,processing and response,so the non-real-time performance of information flow processing directly affects the stability of energy flow and the reliability of control.From the perspective of edge computing and cloud computing collaborative application,this paper first carries out collaborative analysis on edge computing and cloud computing in distribution network scenarios,and then build the"bottom-up,distributed autonomous,cooperative and globally optimized"edge cloud physical distribution information system model,which can faster process business of the distribution network.Finally,deep learning is applied to the model to realize distribution network fault inspection,which verifies the high accuracy of the model.