液体火箭发动机地面试车实时故障检测算法

Real Time Fault Detection Algorithm for Liquid-propellant Rocket Engine Ground Test

为了实时有效地检测液体火箭发动机地面试车中的故障,提出了改进自适应阈值算法(IATA)。讨论了自适应阈值算法(ATA)的原理,研究了其改进方法,提高了算法对故障的敏感性和对新试车数据的适用性。通过大量历史试车数据和实际热试车对IATA算法实时性和有效性的验证表明,IATA算法不仅能够及时地检测到异常试车的故障,同时,对正常试车也没有误报警,简洁有效,计算量小,检测速度快。因此,IATA算法适用于液体火箭发动机地面试车的实时故障检测。

In order to detect liquid rocket engine （LRE） ground test faults in real time effectively, the improved adaptive threshold algorithm （IATA） is proposed. Firstly, the theory of adaptive threshold algorithm （ATA） is discussed, and then the improvement methods are studied to increase the sensitivity for faults and applicability for new tests data. Results verfied with the historical on-fire test data and actual ground test show that the IATA can detect fault and give an alarm in time for the abnormal data, and also have no false alarm for the normal data. The IATA is efficient, easy and fast because it only needs a very little amount of computation. As a conclusion, the IATA is suitable for the real time fault detection of liquid-propellant rocket engines.