An Intelligent aeroengine performance diagnostic method

based on neural networks

was investigated.Sensor measurement deviations from the nominal condition are the only information for the estimation of engine health parameters.They are often distorted by noise and bias

thereby mask the true engine condition and lead to incorrect diagnostic results.A bank of neural networks and engine models was developed.The errors between the model outputs and sensor measurements can be used to detect and isolate the sensor

which has measurement bias.Then the engine performance can be estimated by the set of measurements without biases.The simulation results show that this approach is promising for reliable aero-engine diagnosis.