逼近与跟踪空间翻滚目标的图像视觉伺服控制

During the process of final approaching and tracking tumbling target in space, the visual relative navigation from the chaser to the target may lose efficacy in the dead zone due to the ultra-short distance, thus, only indirect relative navigation information can be used in the position based visual servoing (PBVS), which causes poor control precision. A novel image based visual servoing (IBVS) method is proposed to solve this problem, which directly uses several feature points' coordinates from target's image. The motion and imaging models are firstly established, then functions for the IBVS expected generalized velocities satisfying exponential convergence and the relative position estimation corrections are derived and presented. By introducing these velocities and corrections into the relative attitude and orbit coupled control close-loop based on dual sliding-mode surface control law, an indirect estimation assisted IBVS control method is proposed. Numerical simulations indicate that, the proposed method can dramatically promote the relative attitude and position control accuracies. Hence, it can provide much more safe, accurate, and stable relative attitude and orbit condition for final approaching and tracking the tumbling target in space.