Laboratory Experimentation of Multiple Spacecraft Autonomous Assembly

Abstract

This work introduces a novel approach and its experimental verification for propellant sub-optimal multiple spacecraft assembly via a Linear Quadratic Regulator (LQR). The attitude dynamics of the spacecraft are linearized at each time step, about the current state vector, and the relative dynamics between two spacecraft are assumed as a double integrator. This allows for implementation in real-time of a LQR that computes the optimal gain matrix depending on the current phase of the spacecraft’s mission. As a result, both the attitude and position are sub-optimally controlled. The presented logic compensates for the structural evolution related to an incremental assembly by updating the system’s dynamics matrices. The actuators’ reallocation and command of the assembled structure is dealt with through inter-robot wireless ad-hoc communication. Each spacecraft runs symmetric algorithms, differing only in the number of docking ports that each possesses for the mission, which are related to the number of assembling spacecraft and the final structure’s desired shape. Once the spacecraft are assembled, one acts as master by performing the required navigation and control of the new structure through real-time wireless commanding of the other spacecraft’s actuators. The improved third generation (3G-i) of spacecraft simulators developed at the Spacecraft Robotics Laboratory SRL of the Naval Postgraduate School (NPS) is presented to demonstrate experimental verification of the proposed methodology. Features of the (3G-i) robots include an unique customized construction of rapid prototyped thermoplastic (polycarbonate) that incorporates a lightweight modular design with a small footprint, thus maximizing the entire surface of the SRL robotic testbed.