Machine learning algorithms usage for satellites simulation and control

This work provides an alternative way to control the satellite without using deterministic methods and it shows that simple algorithms show no worse results than more complex approaches in artificial intelligence (such as artificial neural networks, etc.). In this case simple algorithms are better because of their quick prediction, which is important when using these predictions on real artificial satellites. As the result of work, among the proposed simple control solutions, the random forest algorithm proved to be the best, as evidenced by accuracy metrics, as well as presented on the basis of graphs of increasing speed during the simulation