对失控翻滚目标逼近的神经网络自适应滑模控制

针对失控航天器在空间中自由翻滚的情况，研究追踪器对失控翻滚目标逼近的位置和姿态六自由度耦合控制问题。建立追踪器与目标器相对运动的姿轨一体化动力学模型，设计追踪器逼近过程的标称轨迹和标称姿态。综合考虑系统不确定性和外部干扰，设计无抖振的神经网络自适应滑模控制器。将滑模控制与神经网络逼近相结合，采用径向基函数(RBF)神经网络对系统未知部分进行自适应逼近。由Lyapunov方法导出神经网络自适应律，通过自适应权重的调节保证整个闭环系统的稳定性。数值模拟实例说明了所设计的标称轨迹和标称姿态的合理性，同时验证了神经网络自适应滑模控制器的有效性。

Adaptive Sliding Mode Control Based on Neural Network for Approaching to an Uncontrolled Tumbling Satellite

For the situation that an uncontrolled spacecraft is tumbling freely in space, the six-degree-of-freedom coupled control problem of the chaser’s approach to the uncontrolled tumbling target is studied. An integrated dynamic model of attitude and orbit for the relative motion between the chaser and the target is established; besides，the nominal trajectory and nominal attitude of the chaser’s approaching process are designed. With considering the coupled factors of the system uncertainties and external disturbances, an adaptive fuzzy sliding mode controller without chattering is designed. A neural network adaptive sliding mode controller without chattering is designed. The sliding mode control is combined with the neural network approximation, and the unknown part of the system is adaptively approximated by the radical basis function (RBF) neural network. The neural network adaptive law is derived through the Lyapunov method, and the stability of the entire closed-loop system is guaranteed by adjusting the adaptive weights. The numerical simulation example demonstrates the rationality of the designed nominal trajectory and nominal attitude. At the same time, it verifies the effectiveness of the neural network adaptive sliding mode controller.