Minimisation of the Effect of Uncertainty on Model Estimation
U Prells, AW Lees, MI Friswell & MG Smart (University of Wales Swansea)
Mechanical Systems and Signal Processing, Vol. 12, No. 2, March 1998, pp. 333-355
The objective of this paper is to minimise the effect of the uncertain oil film bearing model on the estimation of a foundation model from run-down measurements on a turbo-generator. To do this a method is presented to identify a dynamic transfer matrix relating the response vector of the foundation at the bearings to an estimated force vector which is insensitive with respect to the oil film model. Analytically this transfer matrix is related to that of the foundation simply by the multiplication by a transformation matrix which only depends on the model of the rotor and on the model of the oil film, and which of course retains the sensitivity of the original problem with respect to the oil film model. But since the estimation error of the robust transfer matrix estimate is minimum the retained error is mainly due to the erroneous oil film model used in the transformation matrix. The use of the robust estimate to calculate the contribution of the support on the dynamics of the mounted machine for variations of the oil film model increases the reliability of the predictions.
This material has been published in the Mechanical Systems and Signal Processing, Vol. 12, No. 2, March 1998, pp. 333-355, the only definitive repository of the content that has been certified and accepted after peer review. Copyright and all rights therein are retained by Academic Press. This material may not be copied or reposted without explicit permission.
Please note that Elsevier Science has bought Academic Press. Although the original paper was published on IDEAL, in future access to this journal will be via ScienceDirect.
Download PDF file (300K)
Link to paper using doi:10.1006/mssp.1997.0125
Mechanical Systems and Signal Processing on ScienceDirect