Fault types of sensor of aero-engine is researched.Aiming at characteristic of hard fault

using neural network to train and establish model of aero-engine sensors

according to constructed lines of model

the maximum and minimum threshold value for sensors to work normally is calculated

and the calculated threshold values are embedded to IPC

in which the aero-engine sensors are online real-time simulated on hard fault detection.The algorithm software is written in C + +

which takes little resource but runs fast

high precision and good real-time performance

according to analysis on simulation graph

the program can accurately mornitor fault of aero-engine sensors.