Identification of heat loads on space objects re-entering thee Earth's atmosphere using machine learning methods

The article is devoted to the problem of predicting thermal loads on space objects during reentry.The intensive development of near-Earth space leads to a steady increase in the number of artificial non-functional space objects in Earth orbit, which requires the search for effective methods of space debris removal.Some active and combined methods of space debris removal provide for the possibility of controlling the parameters of re-entry and the impact on the thermal regime of the object.Predicting the amount of heat is important for determining the effectiveness of thermal destruction of an object.The purpose of this work is to assess the possibility of using modern artificial intelligence methods to determine the amount of heat received by a space object while moving in the atmosphere.To solve this task, three popular machine learning algorithms were applied:Support Vector Regression (SVR), Multilayer Perceptron (MLP) and Decision Tree Regression (DTR).Based on a dataset that includes information on entry time, altitude, speed, and air density, machine learning models were trained and compared.Metrics were used to evaluate the prediction accuracy: , mean absolute percentage error, and root mean square error.The results of the study showed that MLP and DTR algorithms demonstrated high forecasting accuracy.MLP and DTR models can be effectively used to predict thermal loads on space objects.In the future, the results obtained can be the basis for creating an effective tool for predicting head load on a space object.