推进技术 ›› 2006, Vol. 27 ›› Issue (2): 141-145.

• • 上一篇    下一篇

涡轮泵故障检测系统

谢光军,胡茑庆   

  1. 国防科技大学机电工程研究所 湖南长沙410073;国防科技大学机电工程研究所 湖南长沙410073
  • 发布日期:2021-08-15
  • 基金资助:
    国家“八六三”高技术研究发展计划资助项目(2002AA722070);国家自然科学基金资助项目(50375153)

Turbopump fault detection system

  1. Inst.of Mechatronic Engineering,National Univ.of Defence Technology,Changsha 410073,China;Inst.of Mechatronic Engineering,National Univ.of Defence Technology,Changsha 410073,China
  • Published:2021-08-15

摘要: 为了提高某大型液体火箭发动机涡轮泵故障检测系统(TP-FDS)的可靠性、实时性和有效性,研究了该系统硬件的构建与布局、多块数据采集卡的同步、以及算法的集成与综合等问题。系统硬件方面,基于工控机构建立了数据记录与存贮子系统(DLSS),基于PX I总线构建了实时故障检测子系统(RT-FDS)和试车后数据分析子系统(PT-DAS);系统软件方面,集成了3种算法来提高实时故障检测的有效性,综合了时域、频域等6方面的算法来保证试车后数据分析的有效性。通过转子试验平台实时数据和液体火箭发动机历史试车数据,验证了TP-FDS。结果表明,该系统具有实时故障检测的能力,其平均故障检测时间低于335m s,故障检测概率达97%,并且能有效地诊断出涡轮泵转子系统的早、中、晚期碰摩故障。

关键词: 液体推进剂火箭发动机;涡轮泵;故障检测

Abstract: In order to improve the reliability,real-time capability and validity of turbopump fault detection system(TP-FDS) for a large liquid rocket engine(LRE),its hardware construction and layout,multi data acquiring cards synchronization,and algorithms integration were studied.As for the TP-FDS hardware,the data logger and saving subsystem(DLSS) was constructed based on industrial computer,while the real-time fault detection subsystem(RT-FDS) and post-test data analysis subsystem(PT-DAS) were developed based on PXI bus;As for the TP-FDS software,3 real time algorithms and 6 types(time domain,frequency domain etc.) post test data analysis algorithms were integrated to enhance its validity of fault detection and data analysis.Then,with the help of historical data from LRE test as well as real-time data from rotor test platform,the TP-FDS was validated.In this case,the TP-FDS has real time fault detection ability,and its mean fault detection time is less than 335ms,while its fault detection probability achieves 97%.At the same time,it can effectively diagnose the early,medium-term,and terminal rub-impact of turbopump rotor.

Key words: Liquid propellant rocket engine;Turbopump;Fault detection