Research on the Co-pyrolysis of Biomass and Coal Based on BP Neural Network

Abstract

This paper takes the statistics of coal pyrolysis data released by the Chinese Academy of Sciences as the research subject and integrates the BP neural network model. It conducts an empirical study on the prediction and optimization issues of product yields during the co-pyrolysis process of biomass and coal. The research reveals that n-hexane insoluble (INS) have an insignificant impact on the yields of tar, water, and char residue. Subsequently, the multi-factor analysis of variance (ANOVA) is employed to analyze the disparities between the experimental values and the theoretical calculated values. The conclusion drawn is that the tar, water, and char residue products exhibit statistically significant differences under certain mixing ratios. Particularly, a notable difference exists between the experimental and theoretical values of char residue, indicating that the existing model might require adjustment and optimization in predicting the yield of char residue. To optimize the yield of co-pyrolysis products, a multivariate regression model is established to analyze the influences of different mixing ratios on the yields of tar, water, and char residue, and the BP neural network model is utilized to predict the yields of pyrolysis products. This showcases the application potential of artificial intelligence in the field of chemical engineering, especially in handling complex chemical reaction systems.