Predictive Health Status Assessment of a Launch Vehicle Engine in an Ascending Flight Based on Vibration Signals

The launch vehicle engine is a typical fuzzy environment in which the health status is extremely difficult to analyze. Current prognostics and health management (PHM) research on rocket engines is almost always a study of real-time health conditions, lacking a significant reference for risk prevention. Moreover, the current evaluation methods are inadequate in terms of universality and are too dependent on expert experience. Our work is the first to propose a multilevel and multifactor predictive evaluation method for the health status assessment of launch vehicle engines. The hierarchy of health assessment is divided by a data-driven method, and the health status of a rocket engine is evaluated based on a prediction algorithm and a fuzzy comprehensive evaluation method. The minimum evaluation error is 0.24% when the method is validated with measured data from long-range launch vehicle engines, which shows that the method presented in this article has a good effect on the prediction and evaluation of launch vehicle engines.