Autonomous Reasoning for Spacecraft Control: A Large Language Model Framework with Group Relative Policy Optimization

2026-01-07

arXiv.org

AI Summary

This work proposes a novel approach that integrates large language models (LLMs) with Group Relative Policy Optimization (GRPO) to address the dual challenges of interpretability and stability in autonomous spacecraft control within complex dynamic environments. By employing a two-stage training process—supervised fine-tuning followed by interactive reinforcement learning—the method introduces an explicit reasoning mechanism into GRPO for the first time, establishing a unified control framework applicable to both linear and nonlinear systems. The framework not only generates stable and feasible control policies, as demonstrated in three-dimensional spacecraft attitude control tasks, but also produces human-readable explanations of its decisions. These capabilities underscore its potential for deployment in safety-critical systems where transparency and reliability are paramount.

Technology Category

Intelligent Robots: Behavior Learning & ControlNatural Language Processing: Safety and RobustnessPlanning, Routing, and Scheduling: Planning with Language Models

Application Category

Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactionsSearch and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved informationResponsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This paper presents a learning-based guidance-and-control approach that couples a reasoning-enabled Large Language Model (LLM) with Group Relative Policy Optimization (GRPO). A two-stage procedure consisting of Supervised Fine-Tuning (SFT) to learn formatting and control primitives, followed by GRPO for interaction-driven policy improvement, trains controllers for each environment. The framework is demonstrated on four control problems spanning a gradient of dynamical complexity, from canonical linear systems through nonlinear oscillatory dynamics to three-dimensional spacecraft attitude control with gyroscopic coupling and thrust constraints. Results demonstrate that an LLM with explicit reasoning, optimized via GRPO, can synthesize feasible stabilizing policies under consistent training settings across both linear and nonlinear systems. The two-stage training methodology enables models to generate control sequences while providing human-readable explanations of their decision-making process. This work establishes a foundation for applying GRPO-based reasoning to autonomous control systems, with potential applications in aerospace and other safety-critical domains.

Problem

• Autonomous Reasoning
• Spacecraft Control
• Large Language Model
• Nonlinear Dynamics
• Safety-Critical Systems

Innovation

• Large Language Model
• Group Relative Policy Optimization
• Autonomous Control
• Reasoning-enabled AI
• Spacecraft Attitude Control