Data-driven Modeling of Nonlinear Joints in Space Structures

The existence of nonlinear joints has a great influence on the dynamics of large flexible sapce structures. As a prerequisite of attitude control and vibration suppression of flexible spacecraft, it is of great importance to establish high-confidence models of joints to predict its dynamic characteristics effectively. First, a single-degree-of-freedom model of the nonlinear joint is established based on the describing function method and harmonic balance method (HBM). In particular, the dynamic funtions of the joints with cubic stiffness and freeplay nonlinearity are derived, and the response and external excitation data are used to train surrogate models by a multi-step DNN method. The validity of the data-driven modeling approach is further verified by response prediction under a different excition frequency. Finally, the characteristic of an unknown nonlinear joint is revealed by the data-driven modeling approach based on the experimental data.