基于混合模型的落叶松微纤丝角模型

以黑龙江省七台河市林业局金沙林场9株人工落叶松2 790个样品数据为例,选择6个常用方程进行非线性回归分析,把拟合精度最高的修正Logistic模型作为微纤丝角基础模型y=b1/1+ exp(b2x)]+b3,然后,利用S-PLUS软件中的NLME过程,拟合非线性微纤丝角模型.采用AIC、BIC、对数似然值和似然比检验等模型评价统计指标对不同模型的精度进行比较分析.结果表明:当对微纤丝角-年龄关系进行拟合时,b1,b2,b3同时作为混合参数时模型拟合效果最好.把相关性结构包括复合对称结构(CS)、一阶自回归结构AR(1)、一阶移动平均结构MA(1)及一阶自回归与移动平均结构ARMA(1,1)]加入到微纤丝角最优混合模型中,一阶自回归与移动平均模型 ARMA(1,1)]显著提高了微纤丝角混合模型的拟合精度.模型检验结果表明:混合模型通过校正随机参数值能提高模型的预测精度.因此,混合模型在应用上不仅能反映总体微纤丝角预测,而且能通过方差协方差结构和误差相关性结构校正随机参数来反映个体微纤丝角差异.

Modeling Microibril Angle with Mixed Models for Dahurian Larch

In this study,the sample data was based on 2 790 samples of 9 trees from dahurian larch(Larix gmelinii) plantations located in Qitaihe Forest Bureau in Heilongjiang Province.The modified Logistic model y=b1/1+exp(b2x)]+b3 was selected to modeling microfibril angle from six models based on nonlinear regression.Then,the logistic model was fitted using nonlinear mixed-effects modeling approach based on NLME of S-PLUS software.Evaluation statistics,such as AIC,BIC,Log Likelihood and Likelihood ratio test were used for model comparisons.The results showed that the Logistic model with parameters b1,b2,b3as mixed effects showed the best performance.Correlation structures included compound-symmetry structure(CS),first-order autoregressive correlation structure AR(1),moving average correlation structure MA(1) and autoregressive-moving average correlation structure were incorprated into the best microfibril angle mixed model.significantly improved the precision of mixed model.Validation confirmed that the mixed model with calibration of random parameters could provide more accurate and precise prediction.Therefore,the application of mixed model not only showed the mean trends of microfibril angle,but also showed the individual difference based on variance-covariance structure and correlation structure.