面向火箭数据分析和故障诊断的数据仓库设计

Design of a Data Warehouse for Rocket Data Analysis and Fault Diagnosis

摘要:

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、显示层六个层次的框架设计；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。

Abstract:

In response to the problem of decentralized management and insufficient utilization of launch vehicle test data, a data warehouse technology is proposed to organize and store the launch data of rockets, achieving deep exploration of the value of rocket data. According to the characteristics of launch vehicle data, six levels of framework design are proposed: data warehouse foundation layer, source data layer, data interaction layer, data architecture layer, application layer, and display layer; Cleaning and loading data from different sources and formats by design data cleaning rules for rocket data analysis and fault diagnosis; Explored the direction of rocket data application and designed data analysis modules such as envelope analysis, correlation analysis, and fault diagnosis. The validation was conducted using launch data from a certain type of rocket, and the results showed that the data warehouse solution was feasible and laid the foundation for the continuous management and analysis of rocket data in the future.