在靶场飞行试验阶段,导弹涡喷(扇)发动机的遥测参数主要有转速、喷嘴前压力等很少几个,飞行结束后供分析的参数是以较长时间轴为坐标的曲线,未经处理的故障特征往往被噪声信号淹没,很难找到高频故障特征.介绍了利用小波变换结合神经网络对遥测数据分析,寻找故障出现的时刻,并结合地面试车故障树数据库,准确定位涡喷发动机的故障类型的方法. In the flight test phase of the shooting range, the telemetry parameters of the missile’s vortex-jet (fan) engine mainly include the number of revolutions and the pressure before the nozzle. The parameters for analysis after the flight are the curves taking the longer time axis as the coordinate and are untreated Is often submerged by noise signals and it is difficult to find the characteristics of high frequency fault.The paper introduces the analysis of telemetry data by using wavelet transform combined with neural network to find out the moment when the fault occurs and combines with the fault tree database of ground commissioning to accurately locate the turbojet engine The type of fault.