Benefits and Strengths of Active AI Pan and Tilt Rotor Control for Unbalanced Drones

A quadcopter drone uses four rotors to maintain flight. A failure of any rotor means the drone will lose balance instantly. Previously, some work was done to allow a single functioning rotor to change the thrust and torque vectors to a new set of angles to attempt to rebalance the drone after a single rotor failure. Another study incorporated an AI controller designed using reinforcement learning to actively change the thrust and torque vectors to reduce the unbalanced effects after a single rotor failure. In this article, a comparison is made between the outcomes of both techniques. The work shows that actively changing the angles of thrust and torque produces better attitude responses on the unbalanced drone compared to passively fixing the angles.