Position–Yaw Tracking of Quadrotors

Abstract

A yaw angle, different from zero, introduces highly nonlinear couplings in the rotational and translational quadrotor dynamics, implying undesirable motions. This argument has motivated that the position control problem of quadrotors is studied generally regulating yaw at zero. However, zeroing yaw limits the maneuverability of underactuated quadrotors because yaw is one of the four actuated motions. In this paper, the simultaneous tracking of position and time-varying heading is proposed, based on an integral sliding mode control with a quaternion-based sliding surface. An exponential tracking with chattering-free is obtained without requiring any knowledge of the dynamic model or its parameters for implementation. Since a linear invariant orientation error manifold is enforced for all time, a time-varying gain is introduced for a well-posed finite time convergence, which is useful not only to realize the virtual position control scheme, due to underactuation, but also to guarantee a desired contact in a given point at a given desired contact time for the yaw motion. Illustrative applications are explored in a simulation study which shows the viability and versatility of position–yaw tracking in the surveillance of a field-of-view (FoV) target, aerial screw driver, and aerial grasping.

Air-arm: A new kind of flying manipulator

“…This can be done by commanding the desired yaw of each multirotor with the inverse kinematics of the flying structure. The yaw control is expected to be realized with recent works about yaw-tracking [1], [83], feasible environment interaction must be researched, and a good starting point is [87], also the aforementioned papers.…”

“…• To develop a suitable mathematical model and the adequate controller; this could be based on [1], [83], [69], [65] investigations. • To develop a way for transmitting force among underactuated flying machines [87].…”

“…Once the master multirotor of the air-arm is placed over the work region, the slave quadrotor just keeps its altitude and yaw with respect to the master multirotor and allows to the master to modify its planar position and altitude, in this way, all the system lets the end effector to perform subjection or any other hand-task. They are needed 3 controllers: 3D position+yaw control for master multirotor; yaw +altitude (not planar) control for slaves multirotors; serial robot manipulator -like controller for commanding all the desired yaw movements through the inverse kinematics of the formation, all these controllers must be modified in order to counteract the variable inertia effects [1], [73], [5], [74], [75], [76], [77], [31].…”

“…The proposed system consist of at least two multirotors and for the purpose of this work, quadrotors linked by rigid bars as if they were a flying serial robot manipulator, where the revolute or rotational joints, instead of being the classical geared motors, are the multirotors by themselves 1 taking advantage of their yaw displacement [1]. UAV manipulation, specifically multirotor manipulation, is a recent line of researching with a boom of at least five years in the field [2], it consists of adding an agent of interaction with the environment to the body of a quadcopter which until recently has been seen just as an article for surveillance and entertaining, in part by the size, in part by the battery performance.…”

“…It is necessary to modify [1] in order to deal with a serial link structure. This article is focused on controlling just one quadcopter but not a chained team 4.…”