摘要: |
以运载火箭贮箱为研究对象,提出了一种基于数据驱动的贮箱故障诊断方法。根据贮箱的结构组成特点和制造流程,分析贮箱制造关键故障模式,建立故障源特征库;通过关键产品特性分析,确定贮箱的故障特征监控点;建立线性回归的故障源检测分析模型,利用线性最小偏差估计方法处理小样本采样数据,根据贡献度和敏感度分析检测故障源;最后针对检测出的故障源进行单值回归预控法分析,实现对故障源的持续监测。 |
关键词: 故障诊断;数据驱动;运载火箭贮箱 |
基金项目: |
|
Study on Fault Diagnosis of Launch Vehicle Tank Based on Data-driven |
Huang Xia1, Chi Hongbo1, Sun Chao1, Sun Ying1, Yu Haidong2, Li Zhimin2, Du Yanfeng1, Tian Zhijie1
|
1. Capital Aerospace Machinery Co., Ltd., Beijing 100076;2. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200241
|
Abstract: |
This paper proposed a data-driven fault diagnosis method for launch vehicle tank. According to the structural characteristics and manufacturing process of the launch vehicle tank, the key failure modes of the tank manufacturing are analyzed, and the feature library of the fault sources is established. By analyzing the key product characteristics, the monitoring points of tank are determined. A linear regression model for fault sources is detected according to the contribution and sensitivity analysis. Finally, the single value regression pre-control method is used to analyze the detected fault sources, and the continuous monitoring of the fault source is realized. |
Key words: fault diagnosis;data-driven;launch vehicle tank |