航天器大规模集群飞行的分布式协调控制方法研究

Space missions of the next generation using large swarm of dozens to hundreds of spacecrafts, such as the sparse aperture observation, and the resource exploration of the asteroid belt, etc., will greatly improve the human abilities of space exploration and exploitation. Coordinated control of swarm motion is the basis for the realization of these missions. The mission life time is bounded by the control performances of fuel consumption and fuel balance. For the coordinated control of large swarm of spacecrafts, size of the problems is huge, and timely global interactions can not be achieved confined by the abilities of sensing, communication and computation of spacecrafts. Thus, current coordinated control methods for formation flying considering only a small number of spacecrafts are difficult to apply. This research is focused on the key coordination control problems of the translational motion of swarm spacecrafts. First, a virtual spring-damper mesh method for rigorous formation keeping is proposed. Free drifting of the swarm as a whole is allowed to reduce fuel consumption. Control forces are adjusted according to the amount of the remaining fuel to achieve fuel balance. Second, a distributed connectivity control based method for loose collective motion maintenance is presented. Only the least constraint of connectivity is reserved to reduce fuel consumption. Fuel balance is considered in the switching and preserving of the links. Third, a distributed planning method for swarm optimal maneuvering based on multidisciplinary design optimization is proposed. Optimal control of swarm maneuvering can be solved distributedly using the computing resources of swarm spacecrafts with low communication cost. The research will provide a theoretical basis for the control of large swarm of spacecrafts, and promote the realization of revolutionary swarm space missions.

下一代包含几十、上百个航天器的大规模集群任务，如稀疏孔径观测、小行星带资源探测等，将极大提升人类探索和利用空间的能力。集群运动协调控制是这类任务实现的基础，其燃料消耗和燃料平衡性能制约了任务寿命。航天器大规模集群飞行，其协调控制问题规模庞大，且受航天器感知、通信和计算能力限制不能及时全局交互，使当前仅考虑少量航天器的编队飞行协调控制方法难以应用。本项目针对航天器大规模集群质心运动的关键协调控制问题，提出严格构形保持的虚拟弹簧阻尼网络方法，容许整体漂移而降低燃耗，按燃料余量调整控制力来实现燃料平衡；提出基于分布式连通性控制的集群松散伴飞维持方法，仅保留连通性这一最少约束来减少燃耗，在连接切换和维持中兼顾燃料平衡；提出基于多学科设计优化的集群最优机动分布式规划方法，能利用集群计算资源以较小通信开销分布式求解最优控制。研究将为航天器大规模集群飞行提供理论与方法基础，促进革命性集群空间任务的实现。

未来革命性的航天任务提出了航天器大规模集群飞行的应用需求。集群运动协调控制是这些任务的基础性技术。针对航天器大规模集群构形控制、集群长期松散相伴飞行维持和构形最优机动规划开展研究。.提出了基于虚拟弹簧阻尼网络的分布式控制方法，以虚拟的弹簧阻尼器连接航天器，以网络连接的虚拟力的合力为依据产生航天器的控制推力。仅依赖航天器间相对运动测量信息和邻居航天器间的局部交互来实现全局构形控制。在虚拟弹簧阻尼网络控制中集成了碰撞避免机制、拓扑结构切换机制，研究了固定构形控制、时变构形跟踪和构形形状控制等问题。分别针对常用的航天器线性相对运动模型和带摄动的引力场非线性模型，利用代数图论、能量方法、切换系统理论等工具从理论上证明了闭环系统的收敛性，导出了稳定性条件，获得了稳态相对运动误差和控制加速度的近似表达式。给出了一种分布式拓扑结构切换准则，使集群能分布式地切换到几何性能更为有利的拓扑连接，并保证切换过程中系统收敛性。理论分析和仿真验证表明，通过增大弹性系数可以使稳态误差充分小，而不会增加控制加速度和燃料消耗。该方法产生的控制力相对集群而言为内力，适用于静电库仑力、电磁力、光压力、系绳张力等基于内力的新型执行机构，可实现无需燃料消耗的构形保持。.开展了航天器集群分布式连通性控制的探索，研究了基于升交点赤经匹配、基于平根数匹配和冲量控制的航天器集群长期相伴飞行的维持方法，从动力学机理上为航天器集群长期相伴飞行提供了依据。.提出了一种航天器集群构形最优机动分布式协同规划方法，采用“分而治之”的求解策略有效降低了问题求解的复杂性，采用响应面代理模型实现顶层规划和底层规划之间的协同，采用重点探索有价值区域的迭代求解方式节省计算开销，通过集群并行计算求解缩短计算时间，为复杂的航天器大规模集群构形最优机动问题提供了求解方案。.项目研究为航天器大规模集群的工程实现提供了运动协调控制的理论与方法基础。