飞机飞行安全依托于航空发动机早期故障诊断。立足于故障诊断信息的收集与融合机理,基于D-S证据理论,从识别框架构建、信任函数指派、故障征兆数据融合和故障模糊诊断四大方面进行了研究,并加以试验。数据分析表明:该方法能够准确诊断航空发动机早期故障并将其控制在萌芽状态,克服了故障征兆和故障原因之间的盘根错节关系,使我国航空技术安全水平迈向了新高度。 Aircraft flight safety relies on early failure diagnosis of aeroengine. Based on the collection and fusion mechanism of fault diagnosis information and based on the theory of D-S evidence, this paper studies and tests four major aspects: identification framework construction, trust function assignment, fault symptom data fusion and fault fuzzy diagnosis. The data analysis shows that this method can accurately diagnose the early failure of aeroengine and control it in its infancy, overcome the faulty relationship between fault symptom and cause of failure, and make the aviation safety in our country reach a new height.