Research Advancements in Artificial Intelligence for Space Situational Awareness

The increasing of satellite launch missions, collisions and disintegration events of space objects, and the vast number of near-Earth asteroids pose significant challenges to space traffic management. Space situational awareness aims to detect, track, classify, and characterize space objects, as well as assess space events, playing a crucial role in space traffic management. The application of artificial intelligence in space situational awareness has unique advantages, especially in enhancing comprehensive abilities such as knowledge, analysis, and decision-making. Its outstanding abilities in transfer learning and lifelong learning can rapidly address complex and dynamic space detection problems. This paper reviews and analyzes the research progress of AI in space situational awareness in the past five years, focusing on data collection, situational awareness, and decision-making. While discussing the advantages and limitations of technical applications, it addresses the feasibility of applying emerging technologies such as distributed machine learning, knowledge graphs, and multi-agent game theory to future space situational awareness research, considering the characteristics of space data and new requirements. This paper points out the effectiveness and necessity of artificial intelligence in space situational awareness, which can provide reference for strengthening space environmental understanding and promoting relevant research in space traffic management.