Espen Oland

Underactuated translational control of a rigid spacecraft

In this paper we consider a spacecraft with one main thruster for translational control and reaction wheels for full attitude control. This is an underactuated control problem, which in this paper is solved using backstepping in order to couple the position tracking problem with the attitude. Assuming that the thrust is non-zero the nonlinear tracking control law is shown to be uniformly globally exponentially stable, and simulations validate these results.

Collision and terrain avoidance for UAVs using the potential field method

In this paper we study the problem of collision and terrain avoidance for multiple Unmanned Aerial Vehicles (UAVs) tracking a low altitude trajectory. By applying the artificial potential field method, the UAVs are able to avoid colliding with the terrain, as well as avoiding collisions with each other. The UAVs are modeled using a kinematic model with nonholonomic contraints, which through dynamic feedback linearization is controllable by using the artificial potential field method.

Reaction wheel design for CubeSats

This paper presents a reaction wheel design for CubeSats where it takes the limitation of size and mass into consideration. It presents an overview of which altitudes it is feasible to use magnetic torquers for momentum dumping as well as presenting equations for customizing reaction wheels for a CubeSat mission. The reaction wheels are then simulated for different CubeSat sizes and proved capable of performing attitude maneuvers. During these simulations a non-linear passivity-based sliding surface controller is used which through Lyapunov stability theory has been shown to be uniformly asymptotically stable.

Subsumption architecture applied to flight control using composite rotations

In this paper the subsumption theory is applied to flight control through composite rotations where multiple tasks can be defined as simple rotations. The tasks can then be arranged as a hierarchy, where the primary task is always fully pursued, and conflicting lower level tasks are removed by the primary rotation. The concept is applied to a group of uavs that move through an urban terrain while avoiding collisions with the ground, the buildings and other uavs as they track a desired waypoint.

Adaptive flight control with constrained actuation

In this paper adaptive flight control in the presence of actuator constraints is considered. Using a quaternion-based approach, the saturation problem can be moved from the plant to the reference trajectory, such that the reference trajectory will diverge from its desired trajectory while the actuators are in saturation. This ensures that the adaptive update laws are not affected by the saturation of the actuators. The underactuation of the aircraft is mapped to a rotation problem using linear filters, which in essence decouples the rotational and translational dynamics such that the controllers can be derived independently of each other. Simulations have been performed where the aircraft tracks a series of waypoints, and it shows good performance even in the presence of unknown aerodynamic forces and moments.

Quaternion-based backstepping control of a fixed-wing unmanned aerial vehicle

In this paper the problem of controlling a fixed-wing UAV is studied. With basis in Newtons second law and Eulers moment equation the translational and rotational dynamics is derived. Using the rotational dynamics a quaternion-based backstepping controller is designed which is able to track a trajectory and is shown to be uniformly asymptotically stable. Since a fixed-wing UAV contains six states and only four actuators, the velocity component in yb and zb directions are underactuated. A velocity controller is designed to make the velocity component in the thrust direction go to zero, while the underactuated states are shown to go to their desired values using a line-of-sight guidance scheme.

A command-filtered backstepping approach to autonomous inspections using a quadrotor

This paper applies command-filtered backstepping for position control of an underactuated quadrotor. First the dynamics are differentiated until the rotational and translational dynamics become coupled, then the paper applies command filtered backstepping to control the underactuated rigid body with constrained actuation. The thrust, roll and pitch torques are used for position control and designed through the backstepping, while the yaw torque is a available for another control objective. The yaw torque is therefore designed using a PD+ controller for target tracking, enabling the quadrotor to track any point (stationary or moving) with a sensor that is assumed to be fixed to the x-axis of the body frame. The paper applies these results to perform autonomous inspections of a wind turbine using a combination of waypoints and a desired trajectory to cover the whole structure while pointing the sensor towards the wind turbine.

Spacecraft formation reconfiguration with dynamic collision avoidance

In this paper we present three different solutions to the collision avoidance problem for spacecraft formations based on the Null-Space Based (NSB) behavioral control concept. In the first case, a constant sized sphere of safety area is centered on each spacecraft and obstacle, and the collision avoidance task is activated when this area is entered using a constant repulsive gain similar to what has been proposed for robotics in earlier publications. In the second case we make use of a variable state dependent gain for increasing the level of repulsiveness for each safety sphere to completely avoid collisions. In the third case the sphere is resized based on the relative position and velocity vectors, such that evasive maneuvers are initiated earlier when on collision course, but kept small when passing by. Through Lyapunov analysis we show that the equilibriumpoint of the follower spacecraft dynamics in closed-loop with a sliding surface based controller is uniformly globally exponentially stable when no collisions are detected, and that collisions will not occur when the collision avoidance task is active by scaling the variable state dependent gain. Simulation results are presented comparing the performance of the proposed methods during a formation reconfiguration maneuver.

Underactuated Waypoint Tracking of a Fixed-Wing UAV*

Abstract In this paper a new method of performing waypoint tracking is shown for underactuated fixed-wing uavs. The position error can be mapped onto the desired axis using a desired rotation matrix, while the velocity error can be mapped to the desired axis using a desired angular velocity. With all errors defined along one axis, the tracking problem is easily solved using only one thruster. A velocity controller is derived which makes sure that the uav tracks a desired total velocity moving towards the next waypoint, while a sliding surface attitude controller is designed to track the desired attitude. The impact of saturation on the attitude controller is also studied where it is shown that the actuators will desaturate in finite time, through a change in the reference trajectory. Using both controllers, a solution to the problem of waypoint tracking of an underactuated uav is proposed, and simulations have been performed that support the theoretical results.

PD+ based spacecraft attitude tracking with magnetometer rate feedback

In this paper we present a solution to spacecraft attitude tracking control utilizing a passivity-based PD+ control solution with magnetometer rate feedback. This solution aims toward small spacecraft with size and weight constraints, which typically carries e.g. solar sensors and magnetometers for attitude determination, but no sensors for angular velocity feedback. The result is a control solution which uses magnetometer rate feedback to provide angular velocity information, while at the same time exploits the natural passivity in the system. The equilibrium points in the resulting closed-loop system are proved to be uniformly asymptotically stable under some mild gain conditions, and controller performance is visualized through simulations.