This paper introduces a method to create engine transient models that retain the fidelity and non-linearity of complex models as well as simplicity and speed of lower fidelity linearized models. The method is based on the design of experiments (DOE) and neural network methodology to create an analytic non-linear model of engine transient operation which has the potential to be used in on-board and off-board applications. The feed forward neural net models were created for a high fidelity model of high bypass turbofan engine (truth model). The performance of the neural net models was verified against the truth model. The verification results showed good agreement between the output of the neural net models and the truth model. Initial investigations also showed a significant reduction in the model execution time.

This content is only available via PDF.
You do not currently have access to this content.