基于RS的森林蓄积量主成分回归估测

Estimation of Forest Volume by Principal Component Regression Based on RS

通过对影响蓄积量的因子进行相关性分析,筛选出与蓄积量存在较好相关性的指标作为自变量。但其自变量间存在多重共线性,会对模型稳定性、预测精度产生影响。通过多元统计分析中的主成分分析法,构造出影响密云县森林蓄积量的主成分,然后与蓄积量进行回归,得到主成分回归,并与一般线性回归模型进行比较。结果表明:主成分线性模型在拟合度、模型适用性与预测精度上都优于一般线性模型。主成分回归模型的复相关系数为0.809,预测精度达到88.26%。

Indexes having a good correlation with stock volume were selected as the independent variables through the analysis of factors affecting stock volume. But there is multicollinearity between the variables which can affect the stability of the mo- del and the accuracy of prediction. The principal component affecting forest reserves in Miyun County were extracted by principal component analysis, and then a prineipal component regression model was obtained using the principal component as independent and forest reserves as dependent. The regression model was compared with the general linear regression model. Results show that the fitting degree, model applicability and prediction accuracy of the principal component regre- ssion model is superior to those of the general linear model. The multiple correlation of the principal component regression model is 0. 809, and the prediction accuracy is 88.26%.