Multi-robot collaborative assembly strategy for large space structure

Using moving robots to assemble space structures is one of the most promising methods for building large spacecraft. This paper aims to reveal the coupling effect among the robots, space environment, and assembly structure, and to provide an optimal multi-robot collaborative assembly strategy, improving assembly accuracy as well as efficiency. First, we construct an assembly dynamic model using the dynamic substructure method, wherein both the vibrations induced by robot movements on the structure surface and the thermal-induced vibrations are considered. In this paper, three-branch robots are adopted to carry out the on-orbit assembly tasks, and the generated force is derived via the Newton-Euler method. The heat load achieved by the space structure is formulated in terms of the relative motions between the sun and the space structures. Then, based on the proposed assembly dynamic model, the vibrations induced by the heat load and the multi-robot movements versus time are deliberately investigated. To improve the accuracy of the assembly structures and assembly efficiency, the genetic algorithm is used to determine the installation position and the path that the robots move from the storage location to the installation one is generated by the A* algorithm and the CBS algorithm, ensuring that (A) global vibrations of space structures are small and do not increase with the increase in the number of robots; (B) computational costs of the assembly process are economic; (C) path conflicts can be avoided when multiple robots conduct assembly task. Finally, numerical experiments validate the advantages of the proposed multi-robot collaborative assembly strategy in the rapid and precise assembly of space structures.