摘要：
By improving the determination method of plastometric indices for the bituminous coal,the tamping plastometric indices (Yd ,Xd )of different blending coals were measured at tamping density of 1.10 t/m3 .Based on multivariate linear regression analysis,the Vdaf-G-Yd coke quality prediction model under tamping density condition was established with the dry ash-free volatile matter content (Vdaf),caking index (G)and tamping maximum thickness of plastic layer (Yd )of different blending coals as independent variables as well as the cold strength indexes (M10 ,M25 )and thermal properties indexes (CRI,CSR)of different cokes as dependent variables.At the same time,the F tests and error analysis were performed on this prediction model.The results show that the tamping plastometric in-dices(Yd,Xd)have a more accurate prediction on the bonding and coking performance of blending coals during the stamp-charging coking.Compared with the traditional Vdaf-G-Y prediction model,the Vdaf-G-Yd prediction model has a better prediction accuracy,and except for the slight increase of rela-tive error of the predicted M10 ,the relative errors of the predicted CRI,CSR and M25 decrease by 5.7%,12.8% and 10.1%,respectively.%通过对烟煤胶质层指数测定方法进行改进,在捣固密度为1.10 t/m3下测定配合煤捣固胶质层指数(Yd ,Xd ),以配合煤的干燥无灰基挥发分(Vdaf )、黏结指数(G)和捣固胶质层最大厚度(Yd )为自变量,分别以焦炭的冷态强度(M10,M25)和热态性能(CRI,CSR)为因变量,利用多元线性回归分析方法,建立捣固密度下的Vdaf-G-Yd 焦炭质量预测模型,并对其进行F检验和误差分析。结果表明,捣固胶质层指数(Yd ,Xd )能更准确地反映配合煤在捣固炼焦过程中的黏结性能与结焦性能；与传统的Vdaf-G-Y焦炭质量预测模型相比,Vdaf-G-Yd 焦炭质量预测模型具有更好的预测精准度,除M10的相对误差略有增加外,CRI、CSR、M25的相对误差均有所减小,其减小幅度分别为5.7%、12.8%和10.1%。
