Adaptive model-free formation-tracking controller and observer for collaborative payload transport by four drones

Abstract

In this paper, a solution is provided for collaborative transportation of a rigid payload by four quadrotors, while a desired formation topology is maintained among the vehicles. Each quadrotor is connected to the payload using a rigid link, where the joints to the payload are located on two perpendicular lines through the center of mass of the payload. The proposed solution includes an adaptive control scheme comprising several modules. Each module is responsible for controlling the motion of a specific subsystem of the entire payload-links-quadrotors system. Since an adaptive model-free control algorithm is utilized for tracking the desired set-points for each module, no information on the inertial parameters of the drones is required. The formation topology is achieved among the quadrotors by a geometry-based solution. Moreover, by utilizing a special linear Kalman filter for estimating the unit vectors along the connecting rigid links, the requirement for position and velocity estimation of the drones is revoked. Instead, the position and velocity of the payload must be estimated by using appropriate sensors. The solution is validated via numerical simulation of transporting a payload along a time-varying trajectory.