Quantitative Structure−Property Relationships for the Estimation of Boiling Point and Flash Point Using a Radial Basis Function Neural Network

Abstract

Radial basis function (RBF) neural network models for the simultaneous estimation of flash point (T f) and boiling point (T b) based on 25 molecular functional groups and their first-order molecular connectivity index (1χ) have been developed. The success of the whole modeling process depended on a network optimization strategy based on biharmonic spline interpolation for the selection of an optimum number of RBF neurons (n) in the hidden layer and their associated spread parameter (σ). The RBF networks were trained by the Orthogonal Least Squares (OLS) learning algorithm. After dividing the total database of 400 compounds into training (134), validation (133), and testing (133), the average absolute errors obtained for the validation and testing sets ranges from 10 °C to 12 °C and 11 °C to14 °C for T f and T b, respectively, and are in agreement with the experimental value of about 10 °C. Results of a standard Partial Least Square (PLS) regression model for single output predictions range from 23 °C to 24 °C and 18 °C to 20 °C for T f and T b, respectively, indicating the superior predictive ability of the neural model and strongly suggests that a nonlinear relationship exists between the input and target parameters of the data. The robustness of the neural model was successfully examined by a random split cross validation based on pooling together of the validation and test data sets. The study shows that the simple numerical coding of a molecule based on its formula together with its 1χ is an attractive way of estimating the flammability properties of organic compounds via an RBF neural network.

Relationship between flash points of some binary base oils and their constituent oils

“…Through versatile equations FP has been related to boiling point [6,[24][25][26], atomic composition [28], and standard energy of vaporization [27]. Quantitative Structure Property Relationship (QSPR) method [7,28] has also been employed for the estimation of FP. In addition, equations have been proposed for the estimation of binary [31], ternary [29], and multi component mixtures [23,27].…”

“…In addition, an initial knowledge about the flash point of mixtures can be regarded as an effective tool in designing the target blends. Several studies have been performed to find the relationship between structure and flash points of organic compounds [5][6][7][8][9][10][11][12] and fuels [13,14]. However, to the best of our knowledge, there is little investigation in the case of blends of lubricating oil [15,16].…”

Artificial Neural Network in Drug Delivery and Pharmaceutical Research

“…The average prediction error for the dielectric constant was 27.0% and for the Kirkwood function 4.1%. Tettech et al have also developed a radial basis forward neural network for simultaneous prediction of flash point and boiling point [94]. The database contained 400 organic compounds with flash points between -60 ºC and 200 ºC.…”

Quantitative Structure–Property Relationship Studies for Predicting Flash Points of Organic Compounds using Support Vector Machines

“…The Average Absolute Error (AAE) for all 400 compounds was 10.3 8C. Afterward, Tetteh et al [4] used a radial basis function neural network for the estimation of flash points for a large set of 400 compounds, and the AAE was 10.2 8C. Katritzky et al [8] studied quantitative structure-flash point relationships for a diverse set of 758 compounds by using Multiple Linear Regression (MLR) and Artificial Neural Network (ANN), and the AAE was 13.9 and 12.6 K, respectively.…”