DECENTRALIZED STATIC SHAPE CONTROL OF LARGE-SCALED GEOMETRY ADAPTIVE TRUSS BY TWO-PERSON GAME APPROACH

摘要

The geometry adaptive truss (GAT) is a structural system with hyper-redundant kinematic degrees of freedom (DOFs) in general and the redundancy is an important feature in the geometry adaptation to its structural mission and/or environment. The static control of geometrical shape becomes a key problem for the geometry adaptation. The increase of the number of kinematic DOFs is beneficial for the flexibility and dexterity of the geometry adaptation, but it makes the size of shape control problem larger at the same time. The decentralization is a promising approach to overcome such a difficulty due to the large problem size. In this study, the shape control problem of large-scaled GAT is reduced to a collection of two-person games defined for the individual of truss modules, and formulated as multi-objective min-max problems in a decentralized manner. Numerical studies are conducted for a backyard structure application and the effectiveness of the proposed approach is demonstrated.