This work presents a low cost system based on Flame Emission Spectroscopy (FES) that enables the prediction of fuel adulteration. The spectral data acquired using FES were associated with chemometric tools--Partial Least Squares Discriminant Analysis (PLS-DA) and Partial Least Squares (PLS), aiming to predict gasoline adulterations with different solvents. The classification of the Brazilian adulterated gasoline samples with turpentine, thinner, kerosene, rubber solvent and ethanol was carried out through a PLS-DA model built using five latent variables (LV) with an accumulated variance of 100% on X and 76.78% on Y. The combination of these techniques provided the discrimination of distinct groups for each one of the studied adulterants. Subsequent to the classification, samples of adulterated gasoline with the same solvents with contents varying from 1 to 50% (v/v) were analyzed through FES and multivariate calibration curves were employed in order to predict the contents of the respective solvents. The results obtained by the combination of FES and PLS provided the determination of gasoline adulterants with small calibration and validation errors and also lower values than the ones reported in the literature using other spectroscopic techniques.
In this work, flame emission spectroscopy (FES) was combined with the partial least squares discriminant analysis (PLS-DA) method aimed at classifying different types of gasoline retailed in gas stations. In Brazil, three different types of gasoline, namely, regular gasoline (RG), gasoline with additives (AG), and premium gasoline (PG), are available for retail. The legislation and literature does not present methods for the discrimination of these types of gasoline and also lacks an agenda with programs that inspect and/or attest to the presence of additives that distinguish these fuels. For each set of samples, spectra were obtained through FES and, subsequently, the results were treated using PLS-DA. The PLS-DA model was built using only three latent variables (LVs) with accumulated variance of 99.98% in X and 51.05% in Y. The model combining FES to PLS-DA provided excellent sensitivity and specificity values for the calibration set and 100% accuracy in predicting. All samples were analyzed as collected at the gas station, and then the results were obtained in a few seconds without any kind of sample preparation.
Recebido em 11/6/12; aceito em 10/9/12; publicado na web em 1/2/2013 Methane combustion was studied by the Westbrook and Dryer model. This well-established simplified mechanism is very useful in combustion science, for computational effort can be notably reduced. In the inversion procedure to be studied, rate constants are obtained from [CO] concentration data. However, when inherent experimental errors in chemical concentrations are considered, an ill-conditioned inverse problem must be solved for which appropriate mathematical algorithms are needed. A recurrent neural network was chosen due to its numerical stability and robustness. The proposed methodology was compared against Simplex and Levenberg-Marquardt, the most used methods for optimization problems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.