【摘要】 以基径和茎长的组合为自变量拟合了杜仲短周期矮林模式中单株叶、皮、杆生物量模型,并进行精度检验.结果表明:单株叶干重和皮干重均以二次函数模型的拟合效果最好;而杆干重以二次项和幂函数模型的拟合效果最好;其中叶、皮、茎干最优模型拟合的方程式分别为Y=2.618+1.097(DL)+0.003(DL)2、Y=0.068+0.387(DL)+0.001(DL)2、Y=25.545+3.042(DL)+0.010(DL)2/Y=0.690(DL)1.377.残差分析结果表明样本数据可靠.模型精度分析表明误差在要求范围内.进一步采用常规的判定系数、估计值标准误和残差分析3个指标来综合评价模型的拟合效果,系统、全面地反映了各模型的优劣.更多还原

【Abstract】 Plant height and basal diameter were used as variables to establish optimal estimation models for leaf,stem,and stalk of Eucommia SRC within one rotation cycle. Results indicated that quadratic term functional model was optimal for leaf and bark biomass evaluations,while both quadratic term functional model and power functional model were most accurate for stalk biomass estamation. The fitted equation was Y = 2.618+ 1.097( DL) + 0.003( DL)2 for leaf biomass estimation,Y = 0. 068 + 0. 387( DL) + 0. 001( DL)2 for bark biomass,and Y = 25.545+ 3.042( DL) + 0.010( DL)2/Y = 0.690( DL)1.377 for stalk biomass,respectively. Samples were verified with residual error to confirm the reliability,and model precision accuracy analysis suggested that errors were within the required range,indicating that models were applicable. Furthermore,coefficient of determination,standard error of the estimate,residual analysis were used to evaluate the fitting effect comprehensively,elaborating merits and demerits of all models.更多还原