Particle Filtering Range Data for Pose Estimation under Torque-Free Motion

S.P. Russell and S.M. Rock (USA)


Robot sensing and data fusion, Pose estimation, Particlefilter, Rendezvous and capture, Simulation, Aerospace


Autonomous rendezvous and capture is a current research interest with varied mission applications ranging from servicing to sample return. Especially interesting is the problem of rendezvous with an uncooperative tumbling target in orbit. One key to this process is the ability to estimate the relative pose and motion of the target object. This paper presents a particle filter method which utilizes a continuous stream of range data to extract the relative pose and rotation rate of the target. This approach circumvents problems encountered by previous techniques, especially the issue of data smearing caused by target motion during data collection. This has implications for estimating faster rotations than previously demonstrated by existing solutions. Furthermore, the proposed method can utilize simpler sensor patterns compared to previous experiments and still correctly estimate the target state. The method is verified via simulation and results are presented.

Important Links:

Go Back