Experimental Evaluation of Vision-Based Automatic Imaging Algorithm for Asteroid Flyby Observation

Flybys, in which a spacecraft approaches a celestial body for exploration, have attracted attention. In flybys, spacecraft trajectory calculations in an offline manner deteriorate the accuracy of tracking imaging due to uncertainties such as geometric parameters during spacecraft navigation. Therefore, vision-based relative trajectory measurement systems are commonly used to improve exploration accuracy. In recent years, however, high-speed flights requiring large maximum angular velocities have been planned, increasing the required target value tracking performance. In such missions, it is difficult to achieve the required performance by simple feedback control due to the effects of camera measurement errors and delay time. To improve the accuracy of flybys, it is effective to design a high-precision online trajectory estimaxtion based on the measured relative trajectories and a high-precision control system for the actuators. This paper proposes a trajectory estimation method based on an online autonomous imaging algorithm. The relative velocity between the spacecraft and the asteroid is estimated using the least-squares method. An accurate target trajectory is generated based on the relative velocity. The drive actuator of the imaging system is controlled to track the generated trajectories. The tracking accuracy of this system was evaluated in experiments using an experimental setup simulating a flyby exploration.