We propose a new position control strategy for VTOL-UAVs using IMU and GPS measurements. Since there is no sensor that measures the attitude, our approach does not rely on the knowledge (or reconstruction) of the system orientation as usually done in the existing literature. Instead, IMU and GPS measurements are directly incorporated in the control law. An important feature of the proposed strategy, is that the accelerometer is used to measure the apparent acceleration of the vehicle, as opposed to only measuring the gravity vector, which would otherwise lead to unexpected performance when the vehicle is accelerating (i.e. not in a hover configuration). Simulation results are provided to demonstrate the performance of the proposed position control strategy in the presence of noise and disturbances.

The existing rigid body attitude controllers (without angular velocity measurements) involve explicitly the attitude in the feedback. Unfortunately, there does not exist any sensor that directly measures the orientation of a rigid body (without any estimation procedure). Therefore the attitude must be generated from the available sensors via some attitude determination (estimation) algorithms. The most recent and efficient attitude estimation algorithms rely on the body vector measurements and the angular velocity (which is assumed to be unavailable in velocity-free attitude controllers). To overcome this circular reasoning-like problem, we propose a velocity-free attitude stabilization control scheme relying solely on body vector measurements. Moreover, the proposed control law is a priori bounded and does not lead to the so-called unwinding phenomenon 1 encountered in some unit-quaternion based attitude control schemes.

Existing position controllers for VTOL UAVs require measurements of the system attitude (orientation) for feedback. However, in practice the orientation cannot be measured directly; it is rather obtained through the use of an attitude observer relying on a set of inertial vector measurements. In this paper a new control strategy is proposed for VTOL-UAVs which avoids the direct measurement of the system attitude (in terms of a rotation matrix, unit-quaternion or other attitude parameterization). In the proposed controller inertial vector measurements are used instead of the system attitude. This eliminates the need for an attitude observer thereby reducing the overall complexity of implementing the closed loop system and avoiding errors that are associated with the attitude observer.

This paper deals with the attitude estimation and control problems for rigid bodies, using inertial vector measurements. First, we revisit the attitude estimation algorithm on SO(3) that has been recently proposed in the literature, and propose some practical extensions and new insightful unit quaternion based proofs. Then, we propose an attitude stabilization control scheme using only inertial vector measurements. The originality of this control strategy stems from the fact that the explicit reconstruction of the attitude as well as the angular velocity measurements are not required anymore.

We propose a new position control scheme for vertical take-off and landing (VTOL) unmanned airborne vehicles (UAVs). Traditionally, the control schemes for this type of vehicle assume that the system attitude is accurately known (or measured). Unfortunately, there does not exist any sensor that directly measures the orientation of a rigid body. Instead, to obtain the orientation of the aircraft, separately designed attitude-estimation schemes relying, for instance, on an inertial measurement unit (IMU) must be employed. Consequently, one drawback of this common practice, on top of the possible inaccuracies in recovering the systems attitude, is mainly related to the difficulty of proving the stability of the overall closed-loop system (observer-controller). Motivated by this problem, we propose a new position control scheme that does not require the recovery of the system's attitude. Instead, we rely on a direct use of the vector measurements provided by the IMU. Our approach can efficiently handle large linear accelerations, which is not the case in traditional controllers relying on IMU-based attitude observers that assume that the accelerometer provides a measurement of the gravity vector in the body-attached frame.

We present two observers that estimate the orientation of a rigid body, which is subjected to linear accelerations and rotational motion, using a global positioning system (GPS) and a body-attached inertial measurement unit (IMU). Unlike some other attitude estimation schemes (which assume that the accelerometer measures the gravity vector, which is not realistic when the rigid body is subject to large linear accelerations), the proposed results belong to the special class of velocity-aided attitude observers, which instead use the true accelerometer measurements (i.e., the system's apparent acceleration). The linear velocity of the rigid body (obtained from the GPS) is used to obviate the requirement of the linear acceleration (which is assumed unavailable in the inertial frame). The new observers can handle large accelerations of the rigid body which could otherwise destroy the performance of other types of attitude observers which assume that the accelerometer measures the gravity vector.

An adaptive position-tracking control scheme is proposed for vertical take-off and landing (VTOL) unmanned airborne vehicles (UAVs) for a set of bounded external disturbances. The control design is achieved in three main steps. The first step is devoted to the design of an a priori bounded linear acceleration driving the translational dynamics toward the desired trajectory. In the second step, we extract the required a priori bounded thrust and the desired attitude, in terms of unit quaternion, from the desired acceleration derived in the first step. In the last step, we design the required torque for the rotational dynamics, allowing the system's attitude to be driven toward the desired attitude obtained at the second step. Two control laws for the system control torque are rigorously designed. The first control law ensures that the position-tracking objective is satisfied for any initial conditions, whereas the second ensures that the tracking objective is satisfied for a set of initial conditions, which is dependant on the control gains. The latter case is included, since it is less complicated than the former control law and may be advantageous from a practical point of view. Finally, simulation results are provided to illustrate the effectiveness of the proposed control strategy.