Model Predictive Control in Spacecraft Rendezvous and Soft Docking

TL;DR

This paper presents a model predictive control approach for spacecraft rendezvous and soft docking, effectively handling nonlinear dynamics through a linear time-varying model and robust pole assignment to ensure precise, oscillation-free docking.

Contribution

It introduces a novel MPC-based control design with robust pole assignment for spacecraft rendezvous, addressing nonlinearities, disturbance rejection, and real-time computation.

Findings

Achieves no overshoot in relative position and attitude

Demonstrates fast online computation

Shows robustness to modeling errors and disturbances

Abstract

This paper discusses translation and attitude control in spacecraft rendezvous and soft docking. The target spacecraft orbit can be either circular or elliptic. The high fidelity model for this problem is intrinsically a nonlinear system but can be viewed as a linear time varying system (LTV). Therefore, a model predictive control (MPC) based design is proposed to deal with the time-varying feature of the problem. A robust pole assignment method is used in the MPC-based design because of the following merits and/or considerations: (a) no overshoot of the relative position and attitude between the target and the chaser to achieve soft docking by placing all closed-loop poles in the negative real axis of the complex plan, which avoids oscillation of the relative position and attitude, in particular, in the final stage, (b) fast on-line computation, (c) modeling error tolerance, and (d) disturbance rejection. We will discuss these considerations and merits, and use a design simulation to demonstrate that the desired performance is indeed achieved.