A TEMPORAL-CAUSAL-GRAPH-BASED FAULT DIAGNOSIS METHOD FOR LIQUID-PROPELLANT ROCKET ENGINES

摘要

Although qualitative model-based methods are simple and quick in the fault diagnosis for complicate systems like liquid-propellant rocket engines (LRE), ambiguity always exists in the final diagnosis results because some important information such as the trend in the change of symptoms, the propagation duration between the nodes in models, the importance of the occurred faults are always neglected by these techniques. So a new diagnosis method for LRE based on the Temporal-Causal-Graph (TCG) is developed in this paper to overcome the above deficiencies. Based on the analysis and clarification of the relation and difference between the TCG and other general causal graphs, the proposed method focuses on the representation of the dynamic behavior of engine operation from the little-deviation dynamical equations. Some techniques including the forward propagation of predicting dynamic behaviors, the backward propagation of generating assumptions of faults, and the progressive monitoring strategy of checking the consistency and producing final diagnosis results are developed. By this way, the generated unreasonable assumptions of faults can be eliminated and the ambiguity of the results can be reduced gradually with the increase of measured information. Simulated results with fault samples of a liquid-propellant rocket engine as example show that the method proposed is effective. It can be available for reference in the design and realization of practical fault diagnosis systems.