Uncertainty propagation in rotor-model-based identification of foundations in rotating machinery

摘要

Identification of foundations in existing rotating machinery installations is still unresolved. A promising strategy, which uses as inputs the rotor model as well as motion measurements of the rotor and the pedestals, is revisited in this paper to identify the bearing pedestal parameters of a rotor bearing foundation system, taking into account the propagation of uncertainties in the input data. A formulation based on multivariate uncertainty analysis is developed and illustrated by applying it to a simple rotor bearing pedestal system with nonlinear bearings. Using a set of pseudo- random-generated repeated inputs, the methodology is examined for two least-squares- based identification algorithms. The results indicate that as the sample size is increased, the reduction in the bias error in the identified parameters is dominated by the data processing strategy rather than the identification algorithm, with minimum bias if the algorithm is applied to input sample means rather than parameter sample means.