Cost of Controls for Multi-Rotor Drones

Abstract This paper focuses on the energy costs of using a control system to follow a specific trajectory as opposed to directly choosing the inputs in an open-loop sense to achieve a specific trajectory. Results show that the PID controller will use 7–13% more energy than the open-loop system for the small-amplitude trajectories and 62–74% more energy for the high-amplitude trajectories while maintaining trajectory errors of approximately 5% and 19%, respectively. The PD controller will use 0.03–0.04% less energy than the open loop system for the small-amplitude trajectories and 0.35% less energy for the high-amplitude trajectories while maintaining similar trajectory errors to the PID controller. The two trajectories have different trajectory errors (5% vs. 19%). Therefore, the controller does not have to move the rotor speeds to the same level, thus causing more energy saved for the large-amplitude trajectory. Based on these simulations, it can be concluded that the PD controller will be significantly more energy efficient than the PID controller while maintaining similar trajectory errors throughout the entire flight. This is because the integral control term in the PID controller accumulates the error so any initial large errors in the position will result in large controller inputs later in the flight. The accumulation of error in the integral term then results in large variations in rotor speed which often leads to a higher energy usage. Thus, a PD controller may be a more energy efficient choice while maintaining effectiveness in trajectory-tracking. These results highlight the importance of considering and comparing the energy costs of various control systems when designing controllers for multi-rotor drones.