Neural Network Model Research of Some Turbofan Engine Based on Recorded Flight Data

Because of the strong non-linearity and time-varying properties, and low precision, bad robustness of the conventional mathematic model, a RBF neural network algorithm of some turbofan engine model was provided. The method made sure of the initial value of RBF neural network parameters and the center of middle layer. In terms of some turbofan engine recorded flight data on a fighter plane, some turbofan engine identification model was set up and the model was based on RBF neural network. The research shows that this method has many advantages such as strong adaptability, fast self-taught ability, high identification precision and good robustness, and is particularly effective for establishing some turbofan engine model