AI-Based Intelligent Power and Propulsion Management for Satellites Using Reinforcement Learning: A Conceptual Proof-of-Concept

This work presents a conceptual reinforcement learning framework for intelligent satellite power and propulsion management. A Q-learning agent learns operational policies that balance power consumption, orbital stability, and battery safety through a multi-objective optimization function in a simulated satellite environment. The project demonstrates the feasibility of AI-assisted decision-making for future autonomous spacecraft operations and serves as an educational proof-of-concept for further research in space systems and artificial intelligence. Project Website: https://vedhiga.github.io/vedhigavb.github.io/