摘要
针对航天器试验任务过程监控的在轨故障诊断状态检测、健康状态评估与航天器寿命预测等多个环节中,海量试验数据在传输、共享、处理、分析、存储等过程中面临的巨大压力,构建面向航天器多源异构数据管理的分布式大数据发布/订阅系统(big-data publish/subscribe system,BPSS)框架。借助大数据与云平台技术,设计了一套航天器海量在轨试验数据传输、管理与缓存方案,对云节点实现主从调控与弹性扩容,并通过基于消息一致性的动态选举算法完成大规模发布/订阅任务,保证数据传输的安全性、一致性与计算效率,具备大数据订阅响应迅速、多源异构数据高吞吐稳定性、分布式组件部署灵活等优点。实地航天器数管实验结果表明,BPSS数据订阅的平均响应时延为0.05 s/GB,同时在单日吞吐量达到85 GB量级时,数据丢帧率控制在0.025%、数据破损率控制在0.018%,与其他开源的发布/订阅系统相比具备一定竞争力。
In various aspects of monitoring the process of spacecraft testing tasks,such as in orbit fault diagnosis state detection,health state assessment,and spacecraft life prediction,massive test data faces enormous pressure in the transmission,sharing,processing,analysis,storage,and other processes.A distributed big-data publish/subscribe system(BPSS)framework for spacecraft multi-source heterogeneous data management is built.With the help of big-data and cloud platform technology,a set of massive in orbit test data transmission,management,and caching solutions for spacecraft is designed.Master-slave regulation and elastic expansion of cloud nodes are implemented,and large-scale publish/subscribe tasks through dynamic election algorithm are completed based on message consistency.It ensures the security,consistency,and computational efficiency of data transmission,and has the advantages of fast big-data subscription response,high throughput stability of multi-source heterogeneous data,and flexible deployment of distributed components.The results of field spacecraft data management experiments show that the average response delay of BPSS data subscription is 0.05 s/GB.At the same time,when the daily throughput reaches the 85 GB level,the data frame loss rate is controlled at 0.025%and the data damage rate is controlled at 0.018%,which has certain competitiveness compared to other open source publish/subscribe systems.
作者
覃润楠
彭晓东
谢文明
惠建江
冯渭春
姜加红
QIN Runnan;PENG Xiaodong;XIE Wenming;HUI Jianjiang;FENG Weichun;JIANG Jiahong(National Space Science Center,Chinese Academy of Sciences,Beijing 100191,China;Unit 63921 of the PLA,Beijing 100094,China)
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2024年第3期963-971,共9页
Systems Engineering and Electronics
关键词
云平台
分布式
大数据
航天器在轨试验
发布/订阅系统
cloud platform
distributed
big-data
spacecraft test in orbit
publish/subscribe system
