针对目前航天器控制系统中计算单元冗余备份过多但系统动态效能不足的问题,提出一种航天器控制系统自组织体系结构,通过高速对等网络将系统建立在云计算的基础上,取消单节点的冗余备份,以系统整体能力为单个节点做备份,当出现节点故障时动态迁移其程序和数据到其他正常节点以继续其功能,从而实现系统在故障时计算体系的动态聚合重构. Aiming at the problem of excessively redundant redundancy of computing unit in spacecraft control system but insufficient dynamic performance of the system, a self-organizing architecture of spacecraft control system is proposed. Based on high-speed P2P network, the system is set up on the basis of cloud computing and canceled Redundant backup of a single node, with the overall capacity of the system as a backup for a single node. When a node failure occurs, it dynamically migrates its programs and data to other normal nodes to continue its function. In this way, the dynamic aggregation refactoring .