Flying Robot Trajectory Tracking Through Metaheuristic-Based Control

Abstract

Flying robots have undergone significant innovations that have profoundly influenced their capabilities and applications. These innovations, especially on flight control systems and artificial intelligence algorithms, had a profound impact on trajectory tracking, revolutionizing the precision and accuracy with which they navigate predefined flight paths. Hence, tracking trajectories accurately presents a significant challenge for flying robots. In this paper, a complete mathematical model of a flying robot known as Quadrotor is presented, which includes in addition to the rotor dynamics, the attitude, and guidance dynamics. Then, control laws based on the Proportional Integral Derivative (PID) controller are designed for trajectory tracking. However, tuning 30 PID gains is a fastidious task and could be time consuming due to the coupling, nonlinearities, and the underactuation of the six degrees of freedom (DOF) system. For that, and in order to enhance the controller performance, Differential Evolution (DE) algorithm is proposed to get the optimal ones, for achieving the desired fast dynamics trajectory. The obtained results show the effectiveness of the designed PID control laws optimized using DE algorithm and the flying robot could perform the reference high frequency spiral trajectory accurately.