脉冲推力下多星协同搬运的预测博弈控制

Predictive game control for on-orbit transportation by multiple microsatellites with impulsive thrust

在轨装配过程中,为了将装配结构运输到装配主体结构附近的期望位置,利用多个微小卫星进行协同搬运。针对多星协同搬运的轨道转移控制问题,为满足多种控制约束的同时减少星载燃料和通信资源消耗,研究了脉冲推力下的多星自主博弈控制方法。考虑到各个微小卫星的独立性和协同性,以相对轨道动力学为模型、以控制精度和能量消耗的二次型为局部目标函数构建多星博弈问题,个体能够通过局部目标函数的优化获得控制策略,避免了传统集中式方法所需的控制分配。鉴于微小卫星的控制能力以及通信能力,引入脉冲推力形式并将其与控制幅值约束合并描述为存在周期及幅值限制的力约束,与避免干扰姿态的力矩约束一起作为多星博弈问题的控制约束,相较于连续推力形式便于实施。在多约束博弈求解方面,设计了预测博弈控制算法来分布式地逼近纳什均衡策略,将计算负担分散在各个微小卫星之间的同时提高了方法的容错性。最后,数值仿真表明:所提出的脉冲推力下的预测博弈控制方法能够在满足多种控制约束条件下使得装配结构达到期望状态,在燃料消耗和通信压力等方面显著优于传统集中式方法。

During on-orbit assembly, multiple microsatellites are employed to transport the assembly structure collaboratively to the desired location near the main assembly structure. To meet several constraints of orbit transfer control during cooperative transportation and reduce consumption of onboard resources, the game-based control method for multiple microsatellites with impulse thrust is studied. Considering the independence and cooperativity of each microsatellite, a game model for cooperative transportation is constructed by taking relative orbital dynamics as the model and the quadratic form of control accuracy and energy consumption as the local objective function. Each microsatellite can optimize the local objective function to obtain the control strategy, avoiding the control allocation required by the traditional centralized method. In view of the control ability and communication ability of microsatellites, the impulse thrust is introduced, which is easier to implement than the continuous thrust. The impulse thrust is combined with the control amplitude constraint as a force constraint with periodic and amplitude restriction, and is taken, together with the torque constraint for avoiding disturbing the attitude, as the control constraints of the game model. To solve the multi-constraint game problem, a predictive game control algorithm is designed to approach the Nash equilibrium strategy in a distributed manner, which distributes the computational burden among microsatellites and improves fault tolerance of the method. Numerical simulation shows that the proposed control method can effectively deal with several control constraints, and is significantly better than the traditional centralized method in terms of fuel consumption and communication pressure.