Subspace and Bootstrap-Based Techniques for Helicopter Model Identification

This paper investigates the problem of helicopter dynamic modeling using a subspace system identification method in conjunction with a bootstrap-based technique for model uncertainty estimation. The paper begins with a brief review of the identification approaches for helicopter dynamic modeling. The MOESP (Multivariable Output-Error State sPace) subspace identification method is then presented and followed by a derivation of a bootstrap-based method for estimating the uncertainty in the identified model. Computer simulations are carried out to illustrate the operation, performance, and effectiveness of the methods using concatenated data sets generated from a small-scale unmanned rotorcraft model. The main contributions of this paper are (i) the combination of a subspace-based method for state-space modeling with a bootstrap-based technique for model uncertainty estimation and (ii) the application of this combination to helicopter dynamic modeling so as to facilitate robust controller design.