驱动航天器姿态的控制系统优化设计

A Optimal Design for Control System of the Attitude of an Underactuated Spacecraft

为使欠驱动刚体航天器由于缺少沿某一轴向的控制输入,而仍能从任意初始姿态完成调节镇定的目标,需要解决姿态系统角速度耦合控制的难题。针对欠驱动航天器姿态角和受耦合控制的角速度难以同时收敛的问题,利用Hopfield神经网络的状态能够自动收敛的特性,引入姿态角作为网络状态,提出了通过优化Hopfield神经网络结构参数来改变网络输出的收敛规律的方法,实现了期望的动力学与运动学系统的同步镇定,并得到其相应控制量。最后利用Lyapunov稳定性理论分析并证明了Hopfield能量函数的全局收敛性。改变网络输出方法不依赖于具体的控制律,具有一定程度上的通用性。仿真结果证明了改变网络输出方法对于欠驱动航天器姿态控制是可行的。

In order to regulate the attitude angle from any initial values, the problem of coupled control with angu- lar rate for attitude angle system needs to be overcome for an underaetuated rigid spacecraft results from the lack of control input in certain axis. Concerning about the problem that attitude angle and angular rate of coupled control are difficult to simultaneously converge for underaetuated spacecraft attitude angle system, an optimal attitude control method for the underaetuated spacecraft is presented. The stabilization of dynamic and kinematics system can be a- chieved synchronously as expected through optimizing the parameters of the net to alter the convergent law of the out- put of the net based on the trait of autonomous convergence of Hopfield neural network＇ s states, and its control value is obtained, which introduces the attitude angle values as the states of the net. Finally, the global convergence of Hopfield energy function is analyzed and demonstrated by Lyapunov stability theory. The method does not depend on any specific control law, with a certain degree of versatility. The numerical simulation results show that the method is feasible to control the attitude of underaetuated spacecraft.