Prediction of Impact Sensitivity of Nonheterocyclic Nitroenergetic Compounds Using Genetic Algorithm and Artificial Neural Network

A quantitative structure-property relationship model was built to predict the impact sensitivity of 186 nonheterocyclic nitroenergetic compounds. The genetic algorithm was employed to select an optimal subset of descriptors that significantly contribute to the impact sensitivity. A nonlinear artificial neural network was employed to fit a possible relationship between the selected descriptors and impact sensitivity. The results are satisfactory for prediction capability, robustness, and generalization. The proposed method can be used to predict the impact sensitivity of nonheterocyclic nitro compounds based on knowledge of the molecular structures.