Study on predicting the stability of penetrating projectile charges via machine learning methods

Abstract

Listen

Translate article

Abstract Due to the limitations of current theoretical research, numerical simulation, and experimental analysis for predicting the stability of penetrating projectile charge, in this work, we propose a method to predict the stability of charges by using machine learning methods. Three models are selected, including Support Vector Machine (SVM), Random Forest (RF), and Extreme Gradient Boosting Tree (XGBoost). A practical sample dataset with inputs of the charge type, temperature and speed, and outputs of whether the charge is stabilized or not is constructed to train the model. The results indicate that using machine learning methods to predict the stability of charges is a feasible research direction.