Woodland quality evaluation with artificial neural network

A growth equation of the stand topping mean height was created by using artificial neural network modeling technology, in Masson pine planted forest. Then, a site index model was created with the created equation as the directed equation, and the site index curved shape was drawn with the method of index level adjustment. Finally, the table of site index was worked out. The model structure is 1:3:1, the total fitting accuracy is 98.01%. Concretely, the mean topping height fitting accuracy of different age is 85.65% to 99.67% and the mean value is 97.92%. The corresponding testing accuracy is 93.88% to 99.87% and the mean value is 97.43%. The result stated clearly that the created model has very high fitting and testing accuracy and very strong prediction ability. The artificial neural network is an effective modeling technology of the stand growth process, and it was proposed to use it extensively in the woodland quality evaluation.