Attitude control with actuator limitation and constraints for LIDAR pointing in close-proximity operations

A monitoring trajectory around an uncooperative space target and the operation of a scanning LIDAR are reproduced by a numerical simulation in MATLAB. At each time step the scanning LIDAR acquires a 3D point cloud that is processed through model-based pose determination algorithms, in order to fully characterize the target-chaser relative motion in terms of relative position and attitude. After that, the estimated pose is used to feed an Unscented Kalman filter whose output is the full relative state, including relative velocity and relative angular velocity information. The relative position measurement and, in particular, the centroid of the measured point cloud can be used to compute the rotation required by the chaser to continuously point the LIDAR boresight axis towards the target. This information allows the control logic (as better explained at section 2.2) to compute the necessary torque to continuously align the boresight axis with the target- chaser relative position.
The simulated relative trajectory is a safety ellipse with a range varying from 25 up to 57 meters and the target rotates with roll (α), pitch (β) and yaw (γ) rate equal to 0.01 deg/s, 0.01 deg/s and 0.4 deg/s, respectively [5]. At the first time step the azimuth and elevation (better defined at section 2.5) representing the line of sight from the LIDAR to the target point centroid, i.e., the initial misalignment of the LIDAR boresight with respect to the target center, are both set to 5 degrees.
A reaction wheels model has been added to this Guidance, Navigation and Control architecture, as said above the commanded torque computed by the controllers is an input to the actuators: thanks to the model it is possible to determine the reaction wheels’ necessary angular velocity to exert that torque (as explained at section 2.4) and that information is also necessary to take friction torque and saturation into account.
To describe relative motion and attitude dynamics it is necessary to specify and define a reference frame, so the parameters we are interested in are expressed in a Chaser body-fixed.
Reference Frame (CRF) which can be simply assumed coincident to the Sensor Reference Frame (SRF) (figure 2.4) since the LIDAR sensor is fixed to the chaser. Particularly it is important to define in this reference frame the two directions to keep aligned: the boresight axis and the line of sight (LoS) of the target.
[...]