Base Reaction Control of Space Manipulators

摘要

In this thesis a research activity is presented, concerning matters of dynamics and control of robot manipulators for space robotics applications. In particular, kinematic control principles suitable for the realization of trajectory-tracking manoeuvres are considered for manipulators in the kinematic redundancy condition, that is typically available on actual space robotic systems. A well known characteristic of space manipulators is due to the dynamic coupling that arises between the manipulator motion and the motion of the base spacecraft on which the manipulator is mounted, due to action-reaction exchanges between the subsystems, and to the characteristics of their momentum equations that determine the non-holonomic nature of the system. This coupling determines dynamic disturbances generated by the manipulator during operation. It reveals important to control these disturbances in order to respect the pointing requirements of the spacecraft, for what concerns communications, on-board instrumentation, and solar panels orientation, and also for what concerns restrictions on the admissible acceleration disturbances exerted on the base platform, that in particular can be imposed during experimental micro-gravity activities onboard the space station. Fundamental objective of the kinematic control schemes developed in this work is to achieve an optimization of the possible joint trajectories that command the movement of the arm, in order to minimize the dynamic disturbances exerted on the platform, which is possible thanks to a balanced coordination of the arm internal motions, that compensates for its momentum variations that are produced during a trajectory-tracking manoeuvre. Original formulations for the base reaction control are presented and analyzed, and the problem is set in the mathematical framework of constrained least squares methods, while the kinematic control is resolved at the joint acceleration level, in order to attain an effective expressions of the kinematic and dynamic variables involved. The proposed principles reveal suitable for real-time space applications, thanks to the local formulation of the optimization problems and to the use of stable and consolidated solution routines. Analysis and validation of the proposed laws have been developed be means of an experimental test campaign on a planar robot manipulator prototype with three degrees-of-freedom, suspended by air bearings on a flat granite plane, in order to simulate the microgravity environment. In particular a series of trajectory-tracking tests have been performed with dynamic measurements of the resultant base reactions. The analysis is completed by means of a robot simulator system, that has been developed by reproducing the geometrical and inertial characteristics of the experimental prototype. The performance of the control laws have been evaluated both in the fixed base condition, and in the free-floating base condition, and in this case an evaluation on the influence of the inertial parameters involved have been carried out. An independent research activity was related to the application of optimization methods for contact forces control of a bio-inspired climbing robots with dry adhesive pads. Control principles are presented, and their performances evaluated by means of a robot simulator and validated through an experimental robot prototype.