eduzhai > Applied Sciences > Engineering >

Developmental Reinforcement Learning of Control Policy of a Quadcopter UAV with Thrust Vectoring Rotors

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: In this paper, we present a novel developmental reinforcement learning-basedcontroller for a quadcopter with thrust vectoring capabilities. This multirotorUAV design has tilt-enabled rotors. It utilizes the rotor force magnitude anddirection to achieve the desired state during flight. The control policy ofthis robot is learned using the policy transfer from the learned controller ofthe quadcopter (comparatively simple UAV design without thrust vectoring). Thisapproach allows learning a control policy for systems with multiple inputs andmultiple outputs. The performance of the learned policy is evaluated byphysics-based simulations for the tasks of hovering and way-point navigation.The flight simulations utilize a flight controller based on reinforcementlearning without any additional PID components. The results show fasterlearning with the presented approach as opposed to learning the control policyfrom scratch for this new UAV design created by modifications in a conventionalquadcopter, i.e., the addition of more degrees of freedom (4-actuators inconventional quadcopter to 8-actuators in tilt-rotor quadcopter). Wedemonstrate the robustness of our learned policy by showing the recovery of thetilt-rotor platform in the simulation from various non-static initialconditions in order to reach a desired state. The developmental policy for thetilt-rotor UAV also showed superior fault tolerance when compared with thepolicy learned from the scratch. The results show the ability of the presentedapproach to bootstrap the learned behavior from a simpler system(lower-dimensional action-space) to a more complex robot (comparativelyhigher-dimensional action-space) and reach better performance faster.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...