Adaptive control of unmanned quadrotor with partial actuator failure using model reference adaptive control (MRAC) with dynamic inversion

Abstract

Unmanned Aerial Vehicles (UAVs) often experience disturbances during operation, which will degrade its performance and may cause potential failure, affecting safety risks to civilians and other 3rd parties. In such event, UAVs with the common cascaded PID control might not be sufficient to compensate for this reduction in performance. This paper proposes to analyze algorithm for adaptive control, mainly focused on the model reference adaptive control (MRAC) such that despite any uncertainty the plant parameters remain close to the behavior of a desired reference model along with dynamic inversion as an adaptive controller for such scenarios which helps in decoupling of the flight states for the non-linear system. The results of the proposed method have been compared with the conventional PID based controller by implanting faults in motor outputs. In addition, the performance of the controls is evaluated in terms of trajectory following capability.