Modeling stand density index based on artificial neural network

A stand density index BP model was created with artificial neural network modeling technology, by taking the average of diameter at breast height as the input variable, and stand number density as the output variable, in Masson pine planted forest. Through training and optimum seeking, the idea model structure is 1:2:1, the fitting accuracy is 98.67%. As a comparison, a Reineke stand density index model was created and fitted with regression analysis method and same sample data, the fitting accuracy is 97.76%. The results comparing with BP model and Reineke model indicate that the artificial neural network is a more effective stand density index modeling technique.