航天器试验大数据系统实现与典型应用

Implementation and Typical Application of Big Date System for Spacecraft Test

摘要:

航天器试验数据具有大数据特征,对于航天器研制中所做的评价和决策具有至关重要的作用。为实现航天器试验数据的高效且有效采集、存储、处理、分析和应用,提高航天器研制水平,分析了航天器试验数据的典型特征；借助大数据技术成果构建了用于航天器试验的大数据系统,描绘了系统部署位置,基于Hadoop、Flume、Kafka、MongoDB、HDFS、Spark Streaming、Presto、Kylin等技术构建了由数据采集层、数据存储层、数据处理层、数据分析层、数据访问层和数据应用层共6层构成的系统架构,阐述了系统各层的主要技术实现途径；探讨了系统在实时监测、健康评估、试验有效性评价、可靠性评估与寿命预测、发现知识、虚拟试验等方面的典型应用与展望。

Abstract:

Spacecraft test data has typical big data characteristic, and is critical to evaluations and decisions made in spacecraft development. In order to achieve efficient and effective acquisition, storage, processing, analysis and application of spacecraft test data, the characteristic of spacecraft test data was analyzed, the big data system for spacecraft test was built, the deployment position was given. The system architecture is based on technology of Hadoop, Flume, Kafka, MongoDB, HDFS, Spark Streaming, Presto, Kylin .etc, and composed of data acquisition layer, data storage layer, data processing layer, data analysis layer, data access layer and data application layer. The implementation approach of each layer was investigated. Typical application and prospect were discussed, such as real-time monitoring, health evaluation, test validity evaluation, reliability assessment and life time forecast, knowledge discovery, virtual test .etc.