摘要
FTU是安装在配电室或馈线上的智能终端设备,可以实现对线路的分段监测和控制。为了在基于FTU业务中获得准确状态,必须保持数据信息的新鲜度,同时考虑极端事件对配电网的影响。该文研究了系统的峰值信息年龄(age of information,AoI),分析了短包传输的块错误概率;利用随机网络演算(stochastic network calculus,SNC)推导峰值AoI及其违背概率;采用A3C(asynchronous advantage actorcritic)方法,最小化峰值AoI违背概率,并对所提算法进行对比验证。发现基于A3C的服务卸载方案可以提高信息新鲜度,在大约1200次事件后节约总时间逐渐平稳,相比于无卸载方案节约总时间最多增加了2.08 ms;相比于块固定方案节约总时间增加了1.03 ms,说明自适应块长度可以有效地降低信息年龄。发现峰值AoI违背概率是FTU的增函数,在100个FTU时,基于A3C的卸载方案的峰值AoI违背概率约为0.12,相比于现有方案的0.27,降低了55.6%。
FTU is an intelligent terminal equipment installed in the distribution room or feeder,which can realize the sectional monitoring and control of the line.In these FTU based services,keeping the freshness of information is very important for accurate control.In order to obtain the accurate state in these applications,it is necessary to maintain the freshness of the data and consider the impact of extreme events on the distribution network.This paper studies the age of information(AOI)of the system.The paper first analyzes the block error probability of short packet transmission,and then deduces the peak signal AOI and its violation probability by using stochastic network calculus(SNC).Finally,the A3C method is used to minimize the peak AOI violation probability,and the proposed algorithm is compared and verified.It is found that the service unloading scheme based on A3C(asynchronous advantage actor critical)can improve the freshness of information.After about 1200 events,the total time saved is gradually stable,and the total time saved is the most compared with the no unloading scheme,an increase of 2.08ms;Compared with the block fixing scheme,the total time saved is increased by 1.03ms,indicating that the adaptive block length can effectively reduce the information age.At the same time,it is found that the peak AOI violation probability is an increasing function of FTU.When there are 100 FTUs,the peak AOI violation probability of A3C based unloading scheme is about 0.12,which is 55.6%lower than 0.27 of the existing scheme.
作者
林文钦
夏炳森
唐元春
张天天
刘依依
LIN Wenqin;XIA Bingsen;TANG Yuanchun;ZHANG Tiantian;LIU Yiyi(Economic and Technological Research Institute,State Grid Fujian Electric Power Co.,Ltd.,Fuzhou 350000,Fujian,China;School of Electrical and Electronic Engineering,North China Electric Power University,Baoding 071000,Hebei,China)
出处
《电网与清洁能源》
CSCD
北大核心
2023年第8期82-88,共7页
Power System and Clean Energy
基金
国网福建省电力有限公司科技项目(SGFJJY00GHJS-2100054)。
关键词
主动配电网
信息采集
A3C
信息年龄
active distribution network
information collection
A3C
AoI