In the maintenance process traditional excessive maintenance of turbofan engines may result in performance deterioration

long maintenance cycle and high maintenance cost. In order to solve this problem effectively

based on turbofan engine overhaul manual and maintenance technology

the fault detection process and repair mode were studied in depth

and the fault diagnosis expert system and fault diagnosis model were established based on BP(back propagation) neural network

verifying the reliability of fault diagnosis model with several sets of real performance data

with the diagnostic accuracy rate up to 95%. Secondly

the two kinds of diagnostic information were combined to develop reliable maintenance program and optimize the maintenance process. Then a maintenance decision method was put forward. By a certain type of turbofan real validation

it shows that this method can effectively eliminate the fault of high exhaust temperature

helping to improve maintenance quality and reduce maintenance cost of the engine.