Prediction of impact sensitivity of nitramine with QSPR method

A quantitative structure-property relationship (QSPR) study was used for prediction of impact sensitivity of nitramine. A set of 35 molecular descriptors were calculated to represent the molecular structures of nitramine. 5 descriptors which have significant contribution to the impact sensitivity were chosen by employing the variable selection method of genetic algorithm(GA). The multiple linear regression(MLR) was then employed to simulate the possible quantitative relationship between these selected descriptors and impact sensitivity. The model was tested by internal and external validations, and the results show good robustness and high prediction ability as well as generalization. The method presented provides a new way for predicting the impact sensitivity of nitramine for engineering.