Constellation Traffic Control Mode Based on Data Mining and Risk Assessment

Due to the explosive growth in the number of spacecraft in orbit worldwide, a large number of international incidents caused by collision and intersection, re-entry and return, space debris mitigation, etc. have emerged, and space traffic management has risen to a national strategic level to ensure the safety of space assets and promote the construction of a strong spacefaring nation. In this study, a mode of constellation traffic control based on data mining and risk assessment was proposed. Firstly, an unsupervised learning data mining clustering method was applied to invert the in-orbit manoeuvre strategy of satellite TLE data and identify constraints on constellation laws. Secondly, the passive constellation configuration constraints, the cluster or formation configuration situation within the constellation and the control box constraint information of the satellite were inversely analyzed to establish a space traffic network operation model of the satellite. Finally, an orbit avoidance Reconfiguration scheme corresponding to the collision warning type was proposed by combining the constellation inversion information. The proposed constellation traffic network control mode was numerically simulated through the constellation of star chains. It was found that the proposed orbit Reconfiguration scheme can greatly decrease the collision risk of space traffic.