Order Reduction of Dynamic Models with Deformation-Conserving Mode Selection and Weighting

To address the excessive model order issue in flexible multibody spacecraft dynamic models developed using the hybrid coordinate method, this paper presents a model order reduction algorithm that achieves dimensionality reduction through selective mode shape prioritization and weighted synthesis. First, a multibody dynamic model for a spacecraft with a central rigid body and flexible appendages is established using the principle of virtual power and the hybrid coordinate method. By introducing nonlinear terms into the deformation description of flexible bodies, the dynamic model becomes applicable to scenarios involving large deformations of flexible components. Then, while constraining the flexible body’s lateral deformation, the model order reduction problem is transformed into an optimization problem with the minimal number of retained modal orders acting as the optimization objective. The proposed methodology achieves model order reduction by employing a data-driven approach to select and weight the dominant modal shapes that predominantly contribute to the dynamic response. Finally, numerical examples validate the effectiveness and generalizability of the proposed reduced-order model compared with the conventional modal truncation model and the full-order model. In addition, an analysis is conducted to investigate the influence of mode shape selection and weighting on the simulation accuracy of the proposed reduced-order model.