基于虚拟滑模控制方法的非合作航天器姿态估计

针对非合作航天器的非线性姿态估计问题,提出一种利用虚拟滑模控制思想实现对目标航天器姿态参数估计的方法。将立体视觉系统输出的实时观测数据作为虚拟控制系统的输入,将航天器的姿态动力学数学模型作为虚拟的控制对象,采用滑模变结构控制器计算出虚拟力矩控制量,从而使虚拟航天器的姿态与观测姿态同步,虚拟航天器姿态即为非合作航天器姿态参数的估计值。仿真实验验证表明,在存在系统误差及状态量初始误差较大的情况下,所提出的基于虚拟滑模控制的估计算法估计效果优于扩展卡尔曼滤波算法,并较好地协调了变结构控制鲁棒性与平滑控制抖振之间的矛盾。

Virtual Sliding Mode Control-based Attitude Estimation for Non-cooperative Spacecraft

A virtual sliding mode control design method is proposed for the nonlinear attitude estimation of non-cooperative spacecraft. The real-time observations outputted by the stereo vision system are used as the inputs of the virtual sliding mode controller, and the spacecraft attitude dynamics mathematical model is taken as a control object of the virtual controller. The virtual control torque is calculated by using the sliding mode variable structure controller. The virtual spacecraft attitude and the observed attitude are synchronous, and the virtual spacecraft attitude is the estimated value of attitude parameter of the non-cooperative spacecraft. Simulation result shows that the proposed estimation algorithm based on virtual sliding mode controller performs better than the extended Kalman filtering algorithm. The application of sliding mode boundary layer control achieves a good balance between the robustness of variable structure control and the smooth control chattering.