西北林学院学报2024，Vol.39Issue(2)：210-216,7.DOI:10.3969/j.issn.1001-7461.2024.02.27
基于混合效应模型的相容性削度-材积方程研究
Compatible Taper-Volume Equation Based on Mixed Effect Model
摘要
Abstract
A system of equations for compatible taper-volume equation system that allows for different defi-nitions of tree volume was developed in this study.Using the fir stem analysis data of Jingouling Forest Farm,Wangqing Forestry Bureau,Jilin Province,the compatible taper-volume equation system was con-structed based on volume ratio equation method and integration method.In fitting model,three parameter estimation methods were used to estimate the parameters of the taper equation,total volume and merchant-able volume equation in both systems,and the random effects of sample trees was considered.The root mean square error and mean absolute residual were selected to compare the accuracy between different sys-tems of equations.Finally,a compatible taper-volume equations system which could accurately predict the diameter of any height and total volume of the stem was developed.The taper equation parameter estima-tion method provided the worst prediction of the merchantable volume and the total volume,and also don't guarantee the best prediction of taper equation.After considered the tree's random effect,the simulation effect was greatly improved whether using the merchantable volume parameter estimation method or the taper equation parameter estimation method.The accuracy of the compatible taper-volume equation system based on the integral method was higher than that of the volume ratio equation method.The compatible ta-per-volume equation system can ensure that the taper equation,the merchantable volume equation and the total volume equation are compatible in form and algorithm.关键词
削度方程/相容性削度-材积方程/非线性似乎不相关回归/混合效应模型/商品材材积Key words
taper equation/compatible taper-volume equation/nonlinear seemingly unrelated regression/mixed effect model/merchantable volume分类
农业科技引用本文复制引用
李春明..基于混合效应模型的相容性削度-材积方程研究[J].西北林学院学报,2024,39(2):210-216,7.基金项目
国家自然科学基金面上项目(31870624). （31870624）
