Control of Quad-rotor in Cooperation with an Attached 3-DOF Manipulator

Abstract

Autonomous quad-rotor flight systems have grabbed considerable attention for varied missions in rescue, inspection devices, and so forth. Accordingly, various control methods have been employed for hovering these devices. Recently, in order to extend the applications of quad-rotors, including, among the others, the pick-and-place, a robotic arm has been attached, which requires the analysis of both dynamic equations and control procedures. However, considering the coupled system as a case of robot-robot interaction, these state-of-the-art flight systems have rarely received attention due to much higher complexity of dynamic equations. Moreover, flight time expense is revealed as another critical issue in the flight systems with the attached arm, which requires effective solution due to excessive arm loads and power source limitations. To this end, in this paper, the dynamic equations of quad-rotor with a 3-link arm are obtained in order to be employed in designing suitable control methods. In this regard, by stabilizing the quad-rotor in desired coordinations and tracking desired paths for robotic arm, the dynamic interaction as the main challenge of the robot-robot interaction is successfully handled. Thereafter, a Linear Quadratic Regulator (LQR) is proposed to tackle the flight time issue in which by assigning optimal values to the state input weighting matrix, the motion of arm links reduced and as the result, the flight time effectively increases. In order to demonstrate the superiority of the proposed method, two well-known control approaches, namely, Sliding Mode Control (SMC) and Pole Placement are implemented in the same conditions. In simulation with Matlab software, the performance of the forgoing methods is compared by employing different indices, where it is inferred that despite presence of an external force resembling windy condition, the proposed LQR decreases the motion index by 24.14% in compare with SMC and Pole Placement methods, with approximately similar tracking index to them.