液氧煤油发动机故障监控特征参数提取研究

(1.西安航天动力研究所,陕西 西安 710100; 2.航天推进技术研究院,陕西 西安 710100)

液氧煤油发动机; 补燃循环; 故障监控; 故障仿真; 特征提取

Research on feature extraction for fault monitoring of LOX/kerosene rocket engine
PU Xingxing1, WANG Jianshe2,GAO Yushan1, LI Chen1, ZHANG Xiaoping1

(1.Xi'an Aerospace Propulsion Institute, Xi'an 710100, China;2.Academy of Aerospace Propulsion Technology, Xi'an 710100, China)

LOX/kerosene rocket engine; staged combustion cycle; fault monitoring; fault simulation; feature extraction

备注

故障监控可以有效提高发动机工作的安全性和可靠性,建立了大推力液氧煤油高压补燃循环发动机稳态故障仿真模型,开展了发动机故障仿真分析,提取了发动机故障监控特征参数。通过对筛选的可能危及发动机安全工作的故障模式进行仿真分析,获得各种故障模式下发动机测量参数对故障的敏感性排序,通过对测量参数相对偏差的相关性分析,获得各测量参数间相关系数和高相关性测量参数组,基于测量参数的敏感性排序和相关性综合提取出发动机故障监控特征参数。

The fault monitoring can improve safety and reliability of rocket engine effectively. A steady fault simulation model of high pressure staged combustion cycle LOX/kerosene rocket engine with high thrust is developed. The fault simulation analysis is conducted and the characteristic parameters of the engine fault monitoring are extracted. The selected fault modes which are possible to endanger the safe running of the engine are simulated and analyzed to obtain the order of engine parameters sensitive to the fault. Based on correlation analysis on relative deviations of measurement parameters, the correlation coefficients among the measurement parameters and measurement parameter sets with high correlation are obtained. The characteristic parameters for fault monitoring of the engine are extracted on the basis of sensitivity order and correlation of measurement parameters.