Abstract: In order to study the effects of three components of biomass, cellulose, hemicellulose and lignin, on biomass molding characteristics, the simplex lattice mixing design method (SLMD) in the mixing experiment method was used to optimize the mixing of three components of biomass samples with different mixing ratios, the mathematical regression model of biomass compact molding performance index was established, and the mathematical regression model was verified by experiments. The results showed that the correlation coefficient of the regression equation of relaxation density and specific energy consumption was more than 0.9998, biomass forming process was not a simple superposition of a single three component compact forming process, there was a certain interaction between the three components in the process of biomass molding, the higher the content of cellulose, the greater the specific energy consumption and the smaller the relaxation density. The higher the lignin content, the smaller the specific energy consumption, but when the lignin content was too high, the energy consumption had a slight upward trend. The higher the content of hemicellulose, the greater the relaxation density, therefore, in order to make biomass had good molding characteristics, that is, relatively low specific energy consumption and relatively high relaxation density, biomass with less cellulose content and higher hemicellulose and lignin contents should be selected as solid fuel. The experimental verification by using cotton straw, bamboo and corn straw showed that the best biomass species for specific energy consumption prediction was corn stalk, and the best biomass species for relaxation density was bamboo, with relative error of 1.54% and 0.8000%, respectively, and the maximum relative error of the regression model of specific energy consumption and relaxation density was 2.64% and 1.0342%, respectively, which indicated that the model had certain prediction effect on actual biomass.

Key words: Simplex-lattice mixed design, Biomass three components, Biomass compaction molding, Specific energy consumption, Relaxation density, Regression model

摘要： 采用单纯形格子混料设计法(SLMD)建立了生物质三组分致密成型性能指标预测模型，并进行了实验验证。结果表明，比能耗和松弛密度回归方程的相关系数均大于0.9998，与实验数据的最大相对误差分别为3.82%和0.5166%。生物质成型过程中三组分对其有一定的交互作用，纤维素比例越高，比能耗越大，松弛密度越小；木质素比例越高，比能耗越小，但比例过高，比能耗有略微上升趋势；半纤维素比例越高，松弛密度越大。用棉花秸秆、毛竹和玉米秸秆进行实验验证，比能耗和松弛密度回归模型的最大相对误差分别为2.64%和1.0342%，表明模型对实际生物质有一定的预测效果。

关键词: 单纯形格子混合设计法, 生物质三组分, 生物质致密成型, 比能耗, 松弛密度, 回归模型