Robust Subspace Identification of a Nonlinear Satellite Using Model Reduction

The frequency-domain nonlinear subspace identification (FNSI) method has recently been successfully applied to large-scale nonlinear structures. One of the key features of FNSI is the nonlinear generalisation of the stabilisation diagram. However, as in linear system identification, the selection of the model order in the diagram is complicated by the presence of spurious poles, resulting from noise and modelling errors. Spurious poles have been shown to strongly perturb the estimation of the nonlinear coefficients. The present paper establishes a constructive procedure to discriminate between spurious and genuine poles. This procedure is derived in modal space and is based on a dominancy index and on model reduction techniques. It is demonstrated on a complete satellite structure possessing nonsmooth nonlinearities and high modal density. Spurious frequency variations in the nonlinear coefficients are proved to be effectively removed, significantly improving the quality of the overall identified model.