There are over 23000 pieces of debris, ranging in size from miniscule particles to larger meteor-sized particles, present in Earth's orbit. Although these particles may appear to be small in size, at high speeds of 15700 mph, they pose catastrophic threats to current operating satellites and future orbital missions. This research study proposes a random forest machine learning classification model that is tasked with accurately predicting the level of risk of various satellite collisions. Through training and testing a model, an accuracy rate of 98.24% is achieved. Additionally, using decision trees, the variables and qualities of satellites are assessed and ranked in terms of determining the final risk of a satellite collision. This research is a suitable asset to modern Space Situational Awareness (SSA), as improved accuracy of predicting risk of space debris collisions is crucial to the future of space debris collision warning systems.
Learned Prediction for Space Debris Collision Risk and Avoidance
2023-11-22
539998 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Autonomous Space Debris Collision Avoidance System
Springer Verlag | 2021
|Orbital debris–debris collision avoidance
Elsevier | 2011
|Orbital debris–debris collision avoidance
Online Contents | 2011
|Space Debris Collision Avoidance for a Space Vehicle
British Library Conference Proceedings | 2012
|