Search

• The geometric configurations of the major component molecules in natural ester insulating oils were constructed. • The kinematic viscosity values of 20 triglycerides at different temperatures were obtained using molecular dynamics. • The molecular descriptors were calculated, and those related to kinematic viscosity were selected and analyzed. • The kinematic viscosity prediction models for triglycerides were constructed based on sparse machine learning methods. The high viscosity property of natural ester insulating oils can be improved by means of molecular structure modification. However, it is difficult to find the optimal molecular structure that meets the conditions in the vast chemical space by relying on macroscopic experiments. Herein, the kinematic viscosity prediction models of triglyceride molecules based on sparse machine learning methods are proposed in this paper. Firstly, the molecular dynamics technique is used to simulate the kinematic viscosity (-20℃ to 40℃) of the main 20 triglyceride molecules in natural ester insulating oil, which provides extensive characterization data for subsequent model training. Secondly, the molecular descriptors are calculated for each triglyceride molecule based on density functional theory (DFT). Thirdly, four sparse feature selection methods (Boruta, RFECV, Pearson correlation coefficient, and mutual information) are used to identify the most relevant molecular descriptors for the kinematic viscosity of natural ester insulating oil and further analysis is performed. Finally, quantitative structure–property relationship (QSPR) models are constructed using automated machine learning methods to achieve efficient and accurate prediction of kinematic viscosity (average R2 reached 0.96). This study provides certain mechanistic explanations from the perspective of molecular descriptors and provides a predictive model basis for the viscosity modification of natural ester molecules, which serves as a theoretical guide to improve the kinematic viscosity of natural ester insulating oil. [ABSTRACT FROM AUTHOR]

• High-performance hybrid thermodynamic model for simulation of molecular separation. • Understanding role of solvent on separation of solutions containing ionic liquid. • Investigation of thermodynamic analysis at the interface using the developed model. Nowadays, industrial-based application of ionic liquids (ILs) has gained numerous attentions due to their brilliant advantages like very low volatility and excellent chemical stability. Various efforts have been recently made to theoretically/experimentally obtain their physicochemical properties and structure–property relationships. The purpose of the current paper is to calculate the surface tension of two prevalently employed ILs including 1-octyl-3-methylimidazolium bis[(trifluoromethyl) sulfonyl] imide and 1-butyl-2,3-dimethylimidazolium bis[(trifluoromethyl)sulfonyl] imide by developing machine learning-based predictive models. To do this, we are dealing with a dataset with many input features. As a result, feature selection plays an important role in our data modeling. Three distinct optimization algorithms were selected to be used for feature selection with the KNN algorithm, including GA, PSO, and FOA optimization. GENETIC-KNN, FOA-KNN, and PSO-KNN had R2 scores of 0.962, 0.941, and 0.827 respectively. With MAPE metric their error rates are 1.67E-02, 2.13E-02, and 4.26E-02 in same order. Finally GENETIC-KNN is selected as the best and novel model in this study. [ABSTRACT FROM AUTHOR]