TECSAS / DEOS

NOTE:

Due to a programmatic reorientation of of the TECSAS project, this activity was stopped in September 2006. Information concerning the subsequent project DEOS (Deutsche Orbitale Servicing Mission) will be presented on our web pages as soon as possible. DEOS will focus on Guidance and Navigation, capturing of non-cooperative as well as cooperative client satellites, performing orbital maneuvers with the coupled system and the controlled de-orbiting of the two coupled satellites. Therefore most of the description given below is still valid for DEOS.

Status of the DEOS Project:

• A Phase-0 was initiated end of April 2007, the Mission Definition Review (MDR) was held end of July 2007.
• A call for Phase-A proposals was released end of February 2008.
• The DEOS Phase-A was finished in February 2009. The study was led by the STI - SpaceTech-i GmbH.
• The DEOS Phase-B will start in January 2010.

A description of the technological goals is given hereafter.

The goal of DEOS is the on-orbit verification of key robotics hard- and software elements for advanced space maintenance and servicing systems. The mission consists of the following phases (Fig. left): far rendezvous, close approach, inspection fly around, formation flight, capture, stabilization and calibration of the compound, compound flight maneuver, active ground control via tele-presence, passive ground control during autonomous operations (monitoring), and controlled de-orbiting of the compound. Robotics is highly involved in capture, stabilization, orbit maneuvers and de-orbiting. The main objective of the capture experiment is to investigate different control strategies and AOCS control modes, as well as to determine suitable maneuvers for soft docking and the subsequent stabilization of the chaser-target spacecraft compound. Two major control strategies are possible for the chaser spacecraft: free-flying or free-floating. In the first case, the AOCS (Attitude and Orientation Control System) intervenes to limit or eliminate any spacecraft motion (therefore the spacecraft can be kept stationary in the operational space), while in the second case the spacecraft is allowed to move in reaction to the robot movements. While the first case (Fig. 1) is more simple to tele-operate and may be necessary to fulfill spacecraft motion constraints (e.g. attitude motion may be limited for communication purposes), the second case (Fig. 2) is more interesting for reducing fuel consumption for spacecraft control and is more safe, since jerky motions arising from thrusters are avoided.

Fig. 1 - Command position (r e) relative to the inertial reference, AOCS compensates disturbances

Fig. 2 - Command position (r e) relative to the inertial reference and allow for base motion. If the spacecraft moves out of the allowed operating window, compensate for it with the spacecraft control.

Capture Position . FY accounts for displacement dY

The chaser will be navigated into the vicinity of the target satellite. A number of images will be taken automatically and dumped to the ground segment. They show the relative position between both spacecrafts and assist to estimate the target motion by the ground operator or by the ground station’s control software. After this phase the chaser moves closer towards the target to have the structure element to be grasped within the workspace of the manipulator.

• In tele-presence mode, the ground operator will position the gripper in front of the structure element by means of stereo video information. After closing the gripper, the compound stabilization takes place.
• In automatic mode, the ground operator selects the structure element to be tracked by means of image processing and enables the automatic capturing, thereafter.

Note that the image processing is done on-ground, thus the vision control loop is closed via the up- and down-link channel (same principle as tested in ROTEX). In automatic mode, it is assumed that the commanded motion of the robot is determined by a motion planning algorithm. This is necessary to verify the feasibility of the grasping maneuver, in terms of robot dynamic (actuator) and geometric (collision avoidance) constraints. The motion planning solution, based on the predictive simulation of the target, provides a joints motion sequence which minimizes some predefined cost criterion (e.g. execution time, spacecraft fuel consumption, preferred grasping configuration for subsequent stabilization).

Dynamic singularities of a free-floating robot are an important issue, too. They are path dependent in operational space, but fixed in joint space, and also dependent on the inertial parameters. If a path planner is used to determine the motions of the robot, the singularities are automatically avoided, since the algorithm works in joint space. However for tele-presence, a supplementary algorithm is necessary to inform the operator if a singularity is being approached. A workspace analysis can be performed to determine the singularity-free workspace, in which the operator can move safely. After closing the gripper, a stabilization motion has to occur to eliminate the residual relative velocity between the two satellites. A good means to damp this motion is to make the robot compliant for the capturing moment and increase the stiffness slowly until a stable compound is achieved. The motion damping in tele-presence mode will also be investigated, where the operator tries to damp the motion by means of tactile and visual feedback.

During orbit- and de-orbiting maneuvers the chaser will push the target rather than pull it, to avoid damage by the thrusters gas exhaust. The chaser thrust force must then not only act in the direction of travel but also through the center of mass of the compound. If this is not so, the whole compound will rotate about its center of mass. The robot will play an active role in order to direct the thrust force correctly through the common center of mass and to achieve the required posture.

Video: showing the capture and stabilization phases Mediaplayer(1.2MB) Realplayer(1.3MB)