| 摘要: |
| 建立ＡＲ模型对液体火箭发动机状态测量参数构成时间序列进行分析，并构造相应的信息距离差别函数进行状态监测。对某次试车的测量参数进行状态监测仿真表明，利用时序分析算法能够在满足实时条件下早期检测出发动机的故障。在算法中，使用了Ｋｏｌｌｂａｃｋ－Ｌｅｉｂｌｅｒ信息距离函数作为判别函数，使得整个计算简单而可靠。 |
| 关键词: 液体推进剂火箭发动机 故障诊断 时序分析 |
| DOI： |
| 分类号:V434 |
| 基金项目: |
|
| A REAL-TIME FAULT DETECTION ALGORITHIMBASED ON TINE SERIES ANALYSIS FOR LRE |
|
Cao Feng1, Cui Dingjun2, Zhang Zhenpeng2
|
|
1.Beijing Fengyuan Machinery INst., Beijing, 100076;2.School of Astronautics, Beijing Univ. of Aeronautics and Astronautics,Beijing, 100083
|
| Abstract: |
| With analysing the time series of measurement parameters of LRE (Liquid Rocket Engine) conditions with AR model, and constructing the relevant function of information distance theconditions of LRE will be monitored’ In this paper, results of simulation for monitoring a practical testdata show that algorithim based on time series analysis can detect fault in real-time conditions. Usingfunction of Kullback--Leibler information distance as the criterion makes the fault detection algorithimeasy and reliable. |
| Key words: Liquid propellant rocket engine Fault diagnosis Time series analysis |
