Visual Pursuit Control with Target Motion Learning via Gaussian Process

In this paper, we propose an observer-based visual pursuit control law which integrates target motion learningvia Gaussian Process (GP). We consider two rigid bodies: a controlled rigid body with a visual sensor, and a targetrigid body whose velocity is unknown. Furthermore, a vision-based motion observer which estimates the target motionis introduced. Then, we propose an enhanced vision-based nonlinear observer and visual pursuit control which employtarget motion learning by GP, where the GP prediction is based on estimated relative rigid body motion. Then, wequantify the performance and prove stability by the notion of uniformly ultimately boundedness. Finally, we demonstratethe effectiveness of the proposed control law through simulations.
Cite as:
J. Yamauchi, T. Beckers, M. Omainska, T. Hatanaka, S. Hirche and M. Fujita, “Visual Pursuit Control with Target Motion Learning via Gaussian Process”, 2020 59th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE), 2020, pp. 1365-1372.