面向火箭数据分析和故障诊断的数据仓库设计

Design of Data Warehouse for Rocket Data Analysis and Fault Diagnosis

针对运载火箭测试发射数据分散存储且数据利用不充分的问题,提出了以数据仓库技术对火箭的测试发射数据进行组织和存储管理,实现对火箭数据价值的深层次发掘;根据运载火箭数据的特点,提出了数据仓库基础层、源数据层、数据交互层、数据架构层、应用层、显示层6个层次的框架设计;通过面向火箭数据分析和故障诊断的数据清洗规则设计,对不同来源、不同格式的数据进行清洗和加载;对火箭的数据应用方向进行了探索,设计了包络分析、相关性分析、故障诊断等数据分析模块;用某型号火箭测试发射数据进行了验证,结果表明,数据仓库方案切实可行,为后续火箭数据的持续性管理和分析奠定了基础。

In response to the problems of decentralized management and insufficient utilization of launch data for rocket testing,a data warehouse technology is proposed to organize and store the test and launch data of rockets,achieving the deep exploration of the rocket data.According to the characteristics of the carrier rocket data,it proposes six levels of framework design such as the data warehouse foundation layer,source data layer,data interaction layer,data architecture layer,application layer,and display layer;The different sources and format data are cleaned and loaded by the data cleaning rule design for the rocket data analysis and fault diagnosis;The direction of rocket data application is explored,which designs data analysis modules such as the envelope analysis,correlation analysis,and fault diagnosis.The validation is conducted using launch data from a certain type of rocket,and the results show that the data warehouse solution is feasible,and it provides a foundation for the continuous management and analysis of the rocket data in the future.