预测木薯茎叶生物量的数学模型

A Mathematical Model on Estimating Stem-leaf Biomass of Cassava

为了构建预测木薯茎叶生物量的数学模型。调查了5个木薯品种茎秆的基部、中部、上部茎粗及茎秆长度和各个分枝的基部粗度、长度等形态学指标,并将生物量分为茎秆和分枝两部分,分别探讨这两部分生物量与相关形态指标的关系。通过多种模型的反复拟合比较,结果以主茎复合幂函数与分枝集合函数的统合方程的拟合效果较好,均方根误差RMSE为104.2,相关系数r为0.9110,拟合优度较高。并且本模型的拟合效果明显优于前人提出的预测木薯茎叶生物量的数学模型。

In order to establish a mathematical model that can forecast the stem and leaf biomass of cassava, several morphological indexes of five cassava varieties were investigated, which included the stem length, the stem diameter in upper, middle and lower parts of cassava, and the length and diameter of each branch. The biomass of stem and branches were studied, and the relationship was analyzed between the two parts and relevant morphological indexes. By comparing a variety of models, it concluded that the integration equation which combined composite power functions on stem and set functions on branches was the best. The RMSE was 104.2, correlation coefficient was 0.9110, and the fitting effect was quite good. And this model was definitely much better than those previously-mentioned mathematical models on forecasting the stem and leaf biomass of cassava.