Multi-Ground Station Collaborative Measurement and Control Technology for Giant Constellation System

Abstract

Measurement and control technology is the key technology to ensure the effi cient operation, maintenance and management of the constellation system.In recent years, with the continuous expansion of the constellation scale, mega-constellation system has gradually formed, which makes the demand for constellation measurement and control show an explosive growth, which puts forward new requirements for the completion of the constellation system measurement and control tasks.Firstly, the constraints of the megaconstellation system measurement and control tasks and the equipment constraints of the ground measurement and control station were analyzed, and the problem modeling was given; Secondly, an interaction method of the ground station agent based on the learning planning segment was proposed, by introduced the constraint penalty operator and the multi-ground station joint the penalty operator was designed to optimized the objective function.Finally, a multi-ground station Agent reinforcement learning algorithm was proposed to solved the multi-ground station cooperative task assignment strategy.Simulation experiments showed that when the task scale was large, the method had a gain of 12%~20% compared with the traditional algorithm in the diff erent scenarios mentioned.

Keywords

• mega constellation;measurement and control task planning;multi-ground station coordination;multi-Agent reinforcement learning