Fast calculation of confidence intervals on parameter estimates of least-squares frequency-domain estimators

Abstract

The least-squares complex frequency-domain (LSCF) estimator—commercially known as the PolyMAX estimator—nowadays is used intensively in various modal analysis applications. The main advantages are the very clear stabilization diagrams and even more important, the speed. In this paper it is shown that confidence intervals of the modal parameter estimates can be derived without major additional calculations, if the frequency response functions are uncorrelated and noise information (e.g. the coherence function) is available. This approach is also applied to the iterative quadratic maximum likelihood (IQML) estimator, indicating strong analogies to the maximum likelihood (ML) estimator. The algorithm is evaluated by means of Monte Carlo simulations.