New Neural Network Group Contribution Model for Estimation of Lower Flammability Limit Temperature of Pure Compounds

In the present study, a group contribution based neural network method is developed to predict the lower flammability limit temperature (LFLT) of pure compounds. The needed parameters of the model are the occurrences of 125 functional groups in every molecule. The average absolute deviation error obtained over 1429 pure compounds used in this study is 2.35%. Therefore, the model is an accurate model and can be used to predict the LFLT of a wide range of pure compounds.