Robust optimal input design for flight vehicle system identification

Test data for flight system identification and parameter estimation are generated by exciting the dynamic systems under investigation with a set of input signals. It is desired to design the inputs such that the information content in the flight test data is maximized, for a given test run time. This can be achieved by applying optimal control methods, if the model structure and preliminary parameter values are known a priori. Most of the optimal input design methods assume ideal conditions. It is assumed that no uncertainty exists in the model structure, a priori parameter values close to the true values are known and an ideal test environment (e.g. no wind) exists. In this paper, a practical approach to design robust optimal inputs is introduced by enhancing the existing solution methods using generalized Polynomial Chaos. The proposed solution method is tested numerically using a simplified nonlinear model of the aircraft longitudinal motion.