CIKAN: Constraint Informed Kolmogorov-Arnold Networks for Autonomous Spacecraft Rendezvous using Time Shift Governor

TL;DR

This paper introduces CIKAN, a novel neural network architecture that incorporates constraints for improved spacecraft rendezvous control, demonstrating superior performance over traditional methods in simulation scenarios.

Contribution

The paper proposes a new CIKAN architecture integrating Kolmogorov-Arnold Networks into constrained neural network approximation for the Time Shift Governor in spacecraft rendezvous.

Findings

CIKAN outperforms MLP-based CINNs in simulations.

Enhanced constraint enforcement in spacecraft rendezvous.

Better trajectory tracking on highly elliptic orbits.

Abstract

The paper considers a Constrained-Informed Neural Network (CINN) approximation for the Time Shift Governor (TSG), which is an add-on scheme to the nominal closed-loop system used to enforce constraints by time-shifting the reference trajectory in spacecraft rendezvous applications. We incorporate Kolmogorov-Arnold Networks (KANs), an emerging architecture in the AI community, as a fundamental component of CINN and propose a Constrained-Informed Kolmogorov-Arnold Network (CIKAN)-based approximation for TSG. We demonstrate the effectiveness of the CIKAN-based TSG through simulations of constrained spacecraft rendezvous missions on highly elliptic orbits and present comparisons between CIKANs, MLP-based CINNs, and the conventional TSG.