Production forecast analysis of BP neural network based on Yimin lignite supercritical water gasification experiment results

Abstract

In the future coal gasification industry, quick and accurate prediction of the gas products can guide industrial production and make production more efficient. This paper carried out the SCWG experiment of Yimin lignite and discussed the effects of temperature, concentration and residence time on gasification. After that, the experimental data were divided into a training set, validation set, and test set according to a ratio of 70%, 15%, and 15%. Then, the regression was performed in the BP neural network, and the number of hidden layers, linear fitting model, and MIV were discussed. The results show that the single-layer neural network has a better fitting effect than the two-layer neural network. The R2 of the ANN model for the products is 0.9921, the RMSE is 0.2952, the MeanRE is 0.0673, and the MaxRE is 0.1957, which is far better than the linear regression. In addition, the mean impact value of temperature, residence time, and concentration is 0.7493, 0.2188, and −0.1051, which shows temperature is the most critical variable.