Speaker
Description
Recent research has witnessed great advancements in the field of flexible landing. Comprising three rigid nodes embedded within a soft body, the flexible lander enhances surface contact area and effectively dissipates residual kinetic energy upon touchdown. Consequently, the flexible lander can mitigate the risk of toppling and rebounding and achieve a more adaptable and reliable surface landing compared with rigid landers. In our previous work, we proposed the convex curvature guidance for small celestial body flexible landing. The convex curvature constraint is formulated and then incorporated into the trajectory optimization performance index using a penalty function approach. By maintaining the trajectory curvature as convex, the flexible lander can enhance early observation of the landing site and alleviate the risk of obstacle collision. Despite its effectiveness, the proposed convex curvature guidance method does not incorporate real-time environment interaction during the landing process; instead, it requires that the trajectory curvature exceeds a predefined threshold. Considering that the threshold value would directly determine the shape of the trajectory and the following landing performance, how to find a proper threshold remains an open question.
For small celestial body exploration missions, sufficient prior information is typically unavailable. In this paper, to enhance the safety of flexible lander during landing in an unknown topography, a collision avoidance guidance with online trajectory curvature adaptation is proposed. To facilitate online adaptation of trajectory curvature, the landing trajectory is projected onto two motion planes, thereby decoupling the three-dimensional curvature constraint into two two-dimensional curvature constraints. Based on this decomposition, a trajectory curvature adaptation method is developed, which adapts the trajectory curvature according to the geometric relationship between the flexible lander’s state and the terrain information. At last, the collision avoidance guidance problem is formulated as an optimal control problem, with the trajectory curvature adaptation encoded as a set of inequality constraints. The effectiveness of the guidance method is verified though a 433 Eros based small celestial body flexible landing numerical simulations. Simulation results demonstrate that the proposed guidance method offers a viable approach for flexible landing guidance, and results in satisfactory performance.