Research On Distributed Coordination Control Of Modular Spacecraft

Recently, with the development of space technology, space systems become very powerful, which causes the complexity of spacecraft’s structure. When some single function of the spacecraft is failure, the whole spacecraft system may be also failure, which will cause a great loss. This motivates the notion of the modular spacecraft which constituted by multiple simple modules. Modular spacecraft has received a lot of attention by countries in the world due to its significant potential in deep space exploration, scientific experiments, ground investigation and military fields. By considering the Lagrange model for the attitude and relative translation dynamics, this dissertation surveys the recent results on multi-agent systems, especially on multiple Largrange systems, and give a deep study on the distributed coordination for the modular spacecraft. The main contents of this dissertation are summarized as follows:The distributed attitude synchronization problem of modular spacecraft system is studied under an undirected graph. We consider three cases: i) distributed attitude consensus without reference, ii) distributed attitude regulation with a fixed reference, and iii)distributed tracking with a varying reference. Control algorithms are proposed based on both state feedback and output feedback, which are based on the module’s own and its neighbors’ information. For the attitude regulation and tracking with a reference, it is assumed that the reference information is obtained by only a subset of the modules and the modules have only local interaction. Passive velocity estimators are designed for the algorithms based on output feedback.The distributed attitude synchronization problem of modular spacecraft system is studied under a general directed graph. We still consider three cases: i) distributed attitude consensus without reference, ii) distributed attitude regulation with a fixed reference, and iii) distributed tracking with a varying reference. Control algorithms are proposed based on both state feedback and the case without relative angular velocity information. We also consider the case with bounded but unknown disturbances, which are restrained by algorithms with adaptive gains. The proposed control algorithms use the module’s own and its neighbors’ information. For the algorithm without using neighbors’ angular velocity information, we show that the closed-loop system is stable if the control gain is chosen large enough. We then propose a control algorithm using gain adaption, which make the proposed algorithm fully distributed.The distributed formation control problem for relative translation of modular spacecraft system is studied under a general directed graph. We consider three cases: i) formation control with full state feedback, ii) formation control without using neighbors’ velocity information, and iii) formation control with gain adaption. We then consider the case with collision avoidance. Control algorithms are proposed based on potential functions. It is also showed that the proposed algorithms can prevent the connectivity maintenance.