Modeling of the Space Shuttle Main Engine Using Feed-forward Neural Networks

Abstract

This paper presents the modeling of the Space Shuttle Main Engine (SSME) using a feed-forward neural network. The input and output data for modeling are obtained from a non-linear performance simulation developed by Rockwell International. The SSME is modeled as a system with two inputs and four outputs. The back-propagation algorithm is used to train the neural network by minimizing the squares of the residuals. The inputs to the network are the delayed values of the selected inputs and outputs of the non-linear simulation. The results obtained from the neural network model are compared with the results obtained from the non-linear simulation. It is shown that a single neural network can be used to model the dynamics of the space shuttle main engine. This neural network model can be used for control design purposes as well as for model-based fault detection studies.