Department of Civil and Environmental Engineering, University of South Carolina, 300 Main Street, Columbia, SC 29208, USA.
Department of Civil and Environmental Engineering, University of South Carolina, 300 Main Street, Columbia, SC 29208, USA.
Bioresour Technol. 2015;187:263-274. doi: 10.1016/j.biortech.2015.03.054. Epub 2015 Mar 17.
The purpose of this study is to develop regression models that describe the role of process conditions and feedstock chemical properties on carbonization product characteristics. Experimental data were collected and compiled from literature-reported carbonization studies and subsequently analyzed using two statistical approaches: multiple linear regression and regression trees. Results from these analyses indicate that both the multiple linear regression and regression tree models fit the product characteristics data well. The regression tree models provide valuable insight into parameter relationships. Relative weight analyses indicate that process conditions are more influential to the solid yields and liquid and gas-phase carbon contents, while feedstock properties are more influential on the hydrochar carbon content, energy content, and the normalized carbon content of the solid.
本研究的目的是开发回归模型，以描述工艺条件和原料化学性质对碳化产物特性的作用。实验数据是从文献报道的碳化研究中收集和整理的，随后使用两种统计方法进行了分析：多元线性回归和回归树。这些分析的结果表明，多元线性回归和回归树模型都很好地拟合了产物特性数据。回归树模型提供了对参数关系的有价值的见解。相对权重分析表明，工艺条件对固体产率和液相及气相碳含量的影响更大，而原料性质对水热炭的碳含量、能量含量以及固体的归一化碳含量的影响更大。
