Model Reduction Techniques and Their Application to Helicopter Models

[摘要] Techniques of model reduction are important not only because a reduced order model may be needed for control system design or for model validation from measured response data, but also because in many applications accurate information about the plant dynamics may only be available, or required, for a restricted range of frequencies. In the content of aircraft flight mechanics models reduced order descriptions are also of considerable importance for handling qualities studies. In this document, various model reduction techniques are reviewed. An 'equivalent system approximation' approach has been selected and applied to the reduction of a helicopter flight dynamics model. The adequacy and degree of accuracy of this 'equivalent system approximation' reduced order model was verified through comparison with a high order model using the Puma helicopter as an example. Excellent agreement between the results from the reduced order model and the original high order system model were obtained over selected range of frequency. Another approximate method --- extended Levy's complex-curve fitting method using a modified least-squares approach has been extended to the multi-input multi-output and has also applied to the reduction of the helicopter flight dynamics model for a Puma helicopter. Very good agreement between the results from the reduced order models and the original system model were again obtained. Comparison of Levy's method with the 'equivalent system' approach showed that in the latter physical insight can be used in the reduction process whereas Levy's method is purely a curve fitting technique. Both techniques can, however, provide useful reduced-order descriptions for given frequency ranges. The extended Levy approach and the 'equivalent systems' approach have both been implemented using the MATLAB software package.

[发布机构] University:University of Glasgow

[关键词] Aerospace engineering