液体火箭发动机振动故障特征信号提取方法
刘英元,陈海峰,耿 直,朱成亮

(西安航天动力试验技术研究所,陕西 西安 710100)

故障诊断; 振动信号; 特征提取; 液体火箭发动机

Extraction method of characteristic signal for vibration fault of liquid rocket engine
LIU Yingyuan, CHEN Haifeng, GENG Zhi, ZHU Chengliang

(Xi'an Aerospace Propulsion Test Technology Institute, Xi'an 710100,China)

fault diagnosis; vibration signal; feature extraction; liquid rocket engine

备注

液体火箭发动机试验是一项高成本、高风险的工程,由于设计缺陷、零件加工误差、工作过程及机械连接结构的影响,导致试验过程发动机大振动现象的发生。基于发动机振动信号分析的状态监测与故障诊断方法是提高发动机可靠性和降低试验成本的重要手段。在给出液体火箭发动机振动故障诊断数学模型的基础上,详细介绍了发动机试验过程中7种特征信号提取方法,即:振幅特征提取、功率谱特征提取、频谱分析(谐频识别和边频识别)、突频特征提取、状态特征提取、小波特征提取和高阶谱特征提取,并结合发动机实际故障诊断方案,给出了发动机特征信号提取算法应用实例,通过发动机热试车验证了液体火箭发动机振动故障诊断的特征信号提取方法的正确性。

The test of liquid rocket engine is a high cost and high risk project, because of the design defects, parts processing errors, working process and mechanical connection structure, which resulting inthe large vibration phenomenon during engine testing.The state monitoring and fault diagnosis method based on engine vibration signal analysis is an important means to improve engine reliability and reduce test cost. A mathematical model for vibration fault diagnosis of liquid rocket engine is presented in this paper. On this basis,seven characteristic signal extraction methods in the process of engine test are introduced in detail. They are amplitude feature extraction,power spectrum feature extraction,spectrum analysis(harmonic frequency identification and edge frequency identification),fundamental frequency feature extraction,state feature extraction,wavelet feature extraction and high-order spectral feature extraction. In addition,combined with the actual fault diagnosis scheme of the engine,an application example of the engine characteristic signal extraction algorithm is given,and this method is verified by the engine hot-fire test.