Prediction of Octane Number of Gasoline Based on Improved Extreme Learning Machine

In order to improve the prediction precision of octane number of gasoline, an improved extreme learning based model is build. The method has a special structure that it has the input neurons not only connected to the hidden neurons but also to the output neurons, which enables the model to explore both the nonlinear and linear components of the inputs-outputs. The standard database of near-infrared-spectrum-based octane number of gasoline is used to verify the validity of the method. The results show that the mean squared errors of the improved extreme learning machine are lower than the extreme learning machine, and R
²
are larger than the extreme learning machine. This means the higher prediction precision of the improved extreme learning method.