Health monitoring of liquid-propellant rocket engines (LRE) is one of the key technologies for improving the safety of existing engines and developing reliable next-generation engines. Extensive research has been done on the health monitoring of the Space Shuttle Main Engine and next-generation reusable LRE. A brief overview of these research projects is presented. Research advances on the health monitoring of the Long March Main Engine YF-20B are described in detail. The failure mode simulation and analysis of the YF-20B engine are introduced. A component module-based diagnosis method is developed, and a fuzzy hypersphere neural network is demonstrated for the fault detection and isolation of the engine. A real-time verie cation system for the health-monitoring algorithms and system was constructed and applied in the research. I. Introduction B ECAUSE of increasingly stringent requirements for the safety, reliability, and operational capabilities of space vehicles and their launch systems in the past 30 years along with the feasibility provided by the progress of modern science and technology, the health-monitoring techniques for liquid-propellant rocket engines (LRE) have undergone signie cant developments. In earlier stages of development, health-monitoring technology was applied in the ground test process for large-scale LRE. In the 1970s, expendable LREs, such as Atlas and Titan, were monitored by redlines, which are limits or thresholds on some important operating parameters, 1 and automatic test data analysis systems were progressively developed for these engines in early 1990s. 2 The partially reusable rocket engine Space Shuttle Main Engine (SSME), developed in the 1970s, is monitored by a condition-monitoring system that is a simple function of redlines. 3 In the 1980s the SAFD (System for Anomaly and Failure Detection ) was developed for SSME ground tests to improve the monitoring performance of the redline system. 1