| This thesis is based on the data obtained by terrestrial laser scanning scan 10 artificial Ko-rean pine forests,a total of 548 trees were combined with the field survey data to construct the Korean pine tree high curve model and the red pine branch high prediction Model and establish simultaneous equations.First,from the five kinds of tree height curve models selected,two mod-els with better fitting effects are selected as the candidate models of the simultaneous equations.Then select one model with good fitting effect and high applicability from the five high-level models under the branches as the basic model,and use the re-parameterization to bring other stand factors into the sub-branch height basic model,and use the optimal subset regression method to select the model with good fitting effect as the sub-branch height alternative model.As a high alternative model.The same method selects the contact high candidate model with good fitting effect.Finally,the tree-height curve model,the under-height candidate model and the contact high alternative model were established in pairs to establish simultaneous equations.Through Seemingly Unrelated Regression Estimation,the best equations are selected by the goodness of fit and the test results,and the simultaneous equations are evaluated.When the op-timal simultaneous equations estimate the tree height,the decision coefficient R~2 is 0.896,and the root mean square error(RMSE)is 0.612m;when the equations estimate the height under the branch,the decision coefficient R~2 is 0.575,mean square The root error(RMSE)is 0.850m;when the equations estimate the contact height,the decision coefficient R~2 is 0.719,and the root mean square error(RMSE)is 0.791m,and various inspection indexes are relatively small.On the whole,the equation system has better fitting accuracy and test effect for tree height,under-branch height and contact height,and can solve the internal correlation problem of tree height,under-branch height and contact height.Dynamic changes provide the foundation. |