A Software Architecture Design for Autonomous Formation Flying Control.
Ming Xu,Yanchao He,Kai Yu
DOI: https://doi.org/10.1109/taes.2017.2721658
IF: 3.491
2017-01-01
IEEE Transactions on Aerospace and Electronic Systems
Abstract:Based on currently developed on-board platforms, this paper proposes a software architecture for orbiting spacecraft for autonomously establishing formation, station-keeping, and reconfiguring maneuvers. Components of the software architecture include orbital prediction (OP), orbital control (OC), and relative navigation (RN); these not only meet the constraining requirements of attitude control, thermal control, thrusters, and the ground-based operator, but also have a minimal impact on the on-board data handling (OBDH) system. All of the OP, OC, and RN algorithms are packaged as a subroutine (OPCNS) called by the OBDH system at regular intervals. The in-advance control instruction set (CIS) is temporarily stored in random access memory to be read by OPCNS as a formal argument, rather than in any fixed space of read only memory. The simplified general perturbations 4 method is introduced to perform a specific time interval of OP based on filtered relative measurements of RN. The relative-orbital-element (ROE) control method is employed to create in-advance instructions, depending on whether or not the predicted ROEs trigger the threshold values of any maneuver and then to allocate these instructions into the CIS following the proposed relationship rules between all of the ROEs and the ground control instruction. To tolerate the temporary divergence of the RN filter after any orbital maneuver, an ROE-based unscented Kalman filter is enhanced by using the control instruction in current execution to accelerate the co- nvergence. A 3-craft formation scenario for interferometric synthetic aperture radar measurements is used to validate all of the functions of the proposed software architecture.